Belder, D.J., Pierson, J.C., Ikin, K., Blanchard, W., Westgate, M.J., Crane, M., Lindenmayer, D.B. (2019) Is bigger always better? Influence of patch attributes on breeding activity of birds in box-gum grassy woodland restoration plantings. Biological Conservation, Vol. 236, pp. 134-152. DOI: https://doi.org/10.1016/j.biocon.2019.05.015 © 2019. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 Is bigger always better? Influence of patch attributes on breeding activity of

2 birds in box-gum grassy woodland restoration plantings

3 *Donna J. BelderAB, Jennifer C. PiersonAC, Karen IkinA, Wade BlanchardA, Martin J.

4 WestgateA, Mason CraneAD, David B. LindenmayerABD

5 A Fenner School of Environment and Society, The Australian National University, Canberra, ACT,

6 Australia 2601.

7 B National Environmental Science Program Threatened Species Recovery Hub, The Australian

8 National University, Canberra, ACT, Australia 2601.

9 C ACT Parks and Conservation, Environment, Planning and Sustainable Development Directorate,

10 ACT Government, Canberra, ACT, Australia 2602.

11 D Sustainable Farms, The Australian National University, Canberra, ACT, Australia 2601.

12 Abstract

13 Restoration plantings are an increasingly common management technique to address habitat loss in

14 agricultural landscapes. Native fauna, including birds, may readily occupy planted areas of

15 vegetation. However, unless restoration plantings support breeding populations, their effectiveness

16 as a conservation strategy may be limited. We assessed breeding activity of birds in box-gum grassy

17 woodland restoration plantings in the South-west Slopes bioregion of New South Wales, Australia.

18 We compared breeding activity in plantings of different size (small and large) and shape (linear and

19 block-shaped) to breeding activity in a set of remnant woodland sites. Contrary to expectations, we

20 found that bird breeding activity was greatest per hectare in small patches. We also found a negative

21 effect of planting age, with younger plantings supporting more breeding activity per hectare. We

22 found no effect of patch type or shape on breeding activity, and that species’ relative abundance

23 was not predictive of their degree of breeding activity. Our results highlight the value of small

24 habitat patches in fragmented agricultural landscapes, and indicate that restoration plantings are as

25 valuable as remnant woodland patches for supporting bird breeding activity. We demonstrate the 1 26 importance of breeding studies for assessing the conservation value of restoration plantings and

27 other habitat patches for avifauna.

28

29 Keywords: Woodland birds, breeding success, SLOSS, restoration, fragmentation, agricultural

30 landscapes

31

1 32 1. Introduction

33 Habitat loss due to land conversion for agriculture is a significant issue globally, with numerous

34 effects on biodiversity and ecosystem processes (Maxwell et al., 2016; Tscharntke et al., 2012).

35 Land clearing is increasing worldwide, particularly in productive agricultural regions (Evans, 2016;

36 Tilman et al., 2017). The extensive removal of native vegetation creates a highly fragmented

37 landscape in which patches of native vegetation exist primarily as small, isolated remnants.

38 Restoration plantings in agricultural landscapes are increasingly implemented to address habitat loss

39 and conserve threatened and declining native fauna, with hundreds of millions of hectares of

40 vegetation being replanted around the world at a cost of billions of dollars (Crouzeilles et al., 2016).

41 To ensure cost-effectiveness and ecological integrity, it is important to quantify the effects of

42 revegetation on biodiversity and assess whether conservation goals are being met, particularly in the

43 long term (Barral et al., 2015; Ruiz-Jaen and Aide, 2005).

44 A core assumption of restoration success is that revegetated patches provide high-quality habitat for

45 the species they are intended to help conserve (Ikin et al., 2016; Ruiz-Jaen and Aide, 2005). In

46 Australia, bird communities that inhabit box-gum grassy woodlands are threatened by ongoing

47 habitat loss and degradation (Rayner et al., 2014), and are a frequent target of restoration efforts

48 (Freudenberger, 2001; Lindenmayer et al., 2013; Smith, 2008). There is evidence suggesting that

49 many bird species will readily occupy restoration plantings, in some cases preferentially inhabiting

50 plantings over remnant woodland patches or other sites (Barrett et al., 2008; Cunningham et al.,

51 2008; Lindenmayer et al., 2016), but how much do we know about the capacity of restoration

52 plantings to support breeding populations of these species? The majority of studies examining avian

53 responses to restoration plantings have used measures such as species richness, diversity, and

54 relative abundance to make inferences about occupancy trends and habitat quality (Belder et al.,

55 2018). However, focusing on occurrence patterns provides a limited picture of how birds are using a

56 site (Chalfoun and Martin, 2007). It is therefore important to quantify whether indicators of long-

57 term persistence, such as breeding activity, follow the same trends. 1

58 1.1. Research objectives

59 The underlying aim of this study was to assess whether birds are able to breed successfully in box-

60 gum grassy woodland restoration plantings. Breeding success can be measured in several ways,

61 with nest success and daily nest survival being commonly-used metrics (Stephens et al., 2004).

62 However, searching for, and monitoring, nests requires considerable time and effort. An alternative,

63 and perhaps more accessible, approach is to use indicators of breeding activity as a proxy for

64 breeding success. For example, a scoring system developed by Mac Nally (2007) ranks

65 observations of breeding behaviour according to how strongly they indicate breeding success (Table

66 1), providing a quantitative measure of the extent to which a given site supports successful breeding

67 (Bennett et al., 2015; Mac Nally et al., 2010; Selwood et al., 2009). A method such as this provides

68 a basis from which to commence the transition from traditional occupancy and abundance surveys

69 to a more population-oriented approach to monitoring avian responses to restoration plantings.

70 Importantly, it also facilitates the collection of breeding data on species of conservation concern,

71 whose nests may be difficult to find in adequate numbers.

72 We sought to investigate bird breeding activity in the context of habitat restoration in a fragmented

73 agricultural landscape. Specifically, we posed the following three questions:

74 Question 1. How does bird breeding activity in restoration plantings compare to breeding activity

75 in remnant woodland patches?

76 We compared breeding activity in restoration plantings, similar-sized woodland remnants, and

77 larger, more intact woodland remnants. In addition to investigating the entire bird assemblage, we

78 assessed breeding activity for species of conservation concern, and cup-nesters vs. dome nesters

79 (Appendix B). Remnant patches are generally considered to be high-value habitat within

80 fragmented agricultural landscapes (Cunningham et al., 2014), and hence we predicted remnant

81 sites would support more breeding activity than restoration plantings. We predicted that breeding

82 activity would be highest in larger woodland remnants than in smaller, more isolated remnants and

1

83 restoration plantings. We made this prediction because comparative studies have shown that species

84 richness and abundance is typically highest in large, intact remnants (Hadley et al., 2018; Helzer

85 and Jelinski, 1999; Martin et al., 2004; Munro et al., 2011). Many species of conservation concern

86 are more closely associated with remnants than plantings (Kinross, 2004), so we also expected to

87 observe more breeding activity from these species in remnants than in plantings.

88 Question 2. How do patch attributes affect breeding activity in plantings and remnant woodland

89 patches?

90 We examined breeding activity in sites of varying size (small and large) and shape (linear and

91 block-shaped). A key finding from pattern-based studies of bird distribution and abundance in

92 fragmented landscapes is that larger patches support more species (Kavanagh et al., 2007; Shanahan

93 et al., 2011; Watson et al., 2003). This is consistent with the resource concentration hypothesis,

94 which posits that there are more resources and thus more individuals and greater species diversity in

95 larger patches (Connor et al., 2007; Root, 1973). Previous species-specific studies have also found

96 that avian reproductive success is positively correlated with patch size (Herkert et al., 2003; Hoover

97 et al., 1995; Luck, 2003; Zanette et al., 2000). We therefore postulated that breeding activity would

98 increase with patch size in parallel with bird species richness and abundance. Similarly, increasing

99 patch linearity is typically associated with lower species richness and abundance (Kinross, 2004;

100 Lindenmayer et al., 2018a, 2007). As such, we predicted more evidence of successful breeding in

101 block-shaped than in linear patches.

102 We predicted a stronger negative response to decreasing patch size and increasing linearity for cup-

103 nesters compared with dome-nesters. This was because edge-effects of predation are stronger in

104 smaller and more linear sites (Fletcher et al. 2007; Helzer and Jelinski, 1999), and cup-nesters tend

105 to be more vulnerable to predation than other nest types (Okada et al., 2017). We also predicted that

106 species of conservation concern, many of which are area-sensitive (Ford et al., 2009; Watson et al.,

107 2005), would show more evidence of breeding activity in larger, block-shaped sites.

1

108 We also tested for an effect of planting age. Previous studies report increases in bird species

109 richness and abundance as plantings mature (Debus et al., 2017; Freeman et al., 2009; Lindenmayer

110 et al., 2016). This is often attributed to the tendency of the vegetation structure and composition of

111 restoration plantings to converge on that of remnant patches over time (Munro et al., 2011). We

112 therefore predicted that increasing planting age would have a positive effect on bird breeding

113 activity.

114 Question 3. Does breeding activity in restoration plantings and remnant woodland patches reflect

115 species assemblage composition?

116 We predicted that breeding activity in our study sites would be reflective of the species assemblage

117 present. That is, we expected the effects of patch attributes (type, size, shape) on relative abundance

118 to be correlated with the effects of patch attributes on breeding activity scores.

119 2. Methods

120 2.1 Study area

121 We conducted this study in the South-west Slopes bioregion of New South Wales, Australia (Figure

122 1). The region is part of Australia’s sheep-wheat belt and has been extensively cleared of native

123 vegetation, with as little as 0.1% of the original vegetation remaining in intact condition (Thiele and

124 Prober, 2000). Remnant patches consist predominantly of white box (Eucalyptus albens) / yellow

125 box (E. melliodora) / Blakely’s red gum (E. blakelyi) grassy woodland, which is a critically-

126 endangered ecological community (Department of the Environment, 2018). Patches of red

127 stringybark (E. macrorhyncha) woodland and mugga ironbark (E. sideroxylon) woodland are also

128 present in our study region.

1

−35.0 Gundagai Wagga Wagga

Vegetation

Type

e

d u

t Reference

i t

a Remnant L −35.5 Planting

NEW SOUTH WALES

−36.0

Albury VICTORIA

146.5 147.0 147.5 148.0 Longitude 129 130 Figure 1 Location of study sites in the South-west Slopes Bioregion of New South Wales, Australia. Map 131 created using ggmap for R (Kahle and Wickham, 2013). 132 133 2.2 Study sites

134 We used spring bird survey data collected over 12 years to select a subset of 12 restoration

135 plantings from a set of long-term monitoring sites (Appendix A) (Cunningham et al. 2007). We

136 selected sites on the basis that they satisfied our criteria for size and shape, and shared at least two

137 of three key species in common – the superb fairywren (Malurus cyaneus), yellow-rumped thornbill

138 (Acanthiza chrysorrhoa), and willie wagtail (Rhipidura leucophrys). We chose these species as they

139 are relatively common, typically found in woodland communities, and encompass the two major 1

140 nest types (one cup-nester and two dome-nesters). Additionally, the yellow-rumped thornbill is a

141 species of conservation concern (Barrett et al., 2003). Nineteen of our 21 chosen sites contained all

142 three target species, with two sites lacking the yellow-rumped thornbill. We attempted to control for

143 the effects of competitive exclusion by selecting sites with low abundances of the noisy miner

144 (Manorina melanocephala), as this hyper-aggressive species is known to have negative impacts on

145 other species of native birds (Bennett et al., 2015; Maron et al., 2013). Our sites were separated

146 geographically by at least 500 m to promote spatial independence.

147 Plantings were aged between 12 and 25 years, 1.3-7.7 ha in area, and 20-200 m in width. A typical

148 planting contained a mature (flowering-age) Eucalyptus overstorey, an Acacia understorey, and a

149 ground layer dominated by annual grasses (both native and exotic). The majority of planted species

150 naturally occur in the study region. Some plantings also contained remnant trees, along with

151 varying amounts of woody debris (fallen trees and branches).

152 We compared plantings with six box-gum grassy woodland remnants, also part of the long-term

153 monitoring study. Remnant patch size ranged from 2.1 to 5.8 ha, with widths of 30-200 m. We also

154 selected three large (47-110 ha) reference sites to represent intact remnant woodland in the study

155 region (two travelling stock reserves, and one remnant on private property). Remnant sites were

156 dominated by a Eucalyptus overstorey, with or without an Acacia understorey, and typically

157 contained woody debris in the form of fallen trees and branches.

158 2.3 Bird surveys

159 To assess breeding activity, we conducted fixed time-per-unit-area surveys (one hour per hectare) in

160 our study sites over two spring breeding seasons. The peak breeding season for the majority of bird

161 species in our study region is September to December (Appendix B). We completed two rounds of

162 surveys in 2015 (October and November), and three rounds in 2016 (September, October,

163 November). We searched sites systematically, identifying and recording indicators of breeding

164 behaviour (Table 1). We designated search areas by the size and shape of sites. For sites < 3 ha, we

1

165 searched 1.3 ha within the site – this was equivalent to the area of the smallest study site. For sites >

166 3 ha, we searched 3 ha within the site. We surveyed block sites in a grid fashion, and linear sites

167 along their length until we had searched the desired area (i.e. 1.3 ha or 3 ha). We surveyed sites

168 throughout the day, with the exception of November 2016 – in this period we completed surveys in

169 the 4 hours post-sunrise and 4 hours pre-sunset. On average, there was an interval of 4.5 weeks

170 between surveys at each site, and we structured the order of site visits to ensure that sites were not

171 consistently surveyed at the same time of day. We did not conduct surveys during inclement

172 weather. All breeding activity surveys were conducted by Author 1.

173 Table 1 Scores allocated to behavioural observations of breeding 174 activity, modified from Mac Nally (2007).

Behaviour Score Feeding of young out of the nest 9.0 Fledglings seen 9.0 Nest with nestlings or feeding of young in the nest 8.0 Presence of juveniles or immature birds 7.5 Fledglings heard 7.5 Adult carrying food 6.0 Nest with eggs or adult on a nest 6.0 Nest empty or under construction (current breeding season) 5.0 Past breeding season’s nest 3.5 Adult gathering nest material 3.0 Courtship 2.0 Territorial behaviour 1.0 Male and female pairs 1.0 175

176 To quantify breeding activity, we used a survey method modified from Mac Nally’s (2007) scoring

177 system. The Mac Nally (2007) method involves calculating an aggregate score of breeding activity

178 in a study site over the course of a study. Scores are calculated based on ranking observations

179 according to how strongly they indicate breeding success (Table 1), with a score of zero indicating

180 no observations of breeding activity. Rather than aggregating breeding activity scores over the

181 course of the study, we modified the method to calculate a score per survey. There were two

182 reasons for this: first, it enabled us to test for effects of factors that may influence detectability of

1

183 bird behaviour during surveys, such as weather and time of day. Second, it enabled us to account

184 for repeat observations of the same individuals or nests across multiple surveys.

185 We conducted point count surveys in 2016 to quantify bird community composition and abundance

186 in our study sites. Point count surveys in each site were typically conducted within two days of the

187 surveys for breeding activity, and usually on the same day. Point count surveys in September were

188 conducted by Author 1, and in October and November were completed by different observers (the

189 entirety of each month’s surveys conducted by a different observer). We divided each study site

190 into 25 x 25 m cells, and randomly selected cells in which to conduct point counts. For sites > 3 ha,

191 we selected six cells, and for sites <3 ha, we selected three cells. We ensured adjacent cells were not

192 selected. At the centre of each randomly-chosen cell, we completed a five-minute count, recording

193 counts of birds detected within 50 m of the survey point.

194 2.4 Statistical analyses

195 We used a model selection approach to investigate the effects of patch attributes on the total

196 breeding activity score recorded in each survey (Table 2). We used linear mixed effects regression

197 models with study site and survey year as random effects to account for repeated visits to sites over

198 multiple years. The explanatory variables of primary interest were site type, size, and shape, and

199 age of plantings. We included the variable “fenced”, to account for potential effects of cattle

200 grazing in our study sites (Lindenmayer et al., 2018b). Our response variable was a total breeding

201 activity score standardised by survey area (1.3 or 3.0 ha), and was square-root-transformed to

202 improve the distribution of the data. We also scaled and centred our continuous predictor variables.

203 Prior to fitting models with our explanatory variables of interest, we examined variables likely to

204 influence detectability in surveys, including time of day, temperature, and wind. In addition, we

205 accounted for variation in activity through the breeding season by including Julian date. We found

206 that breeding activity increased with Julian date for the woodland assemblage and all subsets of the

1

207 assemblage, so included it as an explanatory variable in subsequent models. There were no other

208 weather or temporal variables of statistical significance (Appendix D).

209 Prior to fitting models, we checked all explanatory variables for multi-collinearity using variance

210 inflation factors. We corrected for multi-collinearity by removing large reference sites from models

211 that included both size and shape. We also removed temperature due to its correlation (0.53) with

212 time of day. We checked for a quadratic effect of time of day and found none. After fitting models,

213 we checked for spatial autocorrelation in the data using variograms of the residuals. We detected no

214 evidence of a nugget or sill in the variograms, and therefore assumed no spatial autocorrelation.

215 For our analyses, we included data for all terrestrial species recorded during breeding activity

216 surveys, with the exception of introduced species (

1

217 Appendix B). We hereafter refer to this assemblage as the “woodland assemblage”. For babblers

218 and finches, we included data on nests only when they could be positively identified as true nests –

219 these species build roost nests, which can be difficult to distinguish from true nests. We subset the

220 woodland assemblage to investigate species of conservation concern, and compared cup-nesters

221 with dome-nesters. We defined species of conservation concern as those listed as threatened in New

222 South Wales (NSW Environment and Heritage 2018), along with those whose reporting rates

223 declined by >20% in the South-west Slopes bioregion between the first and second Atlas of

224 Australian Birds (Barrett et al. 2003). We classified cup-nesters and dome-nesters as per Morcombe

225 (2003) and Pizzey and Knight (1997). The dome-nester group was highly correlated (0.79) with the

226 woodland assemblage, as were species of least concern (0.91), so we did not analyse these groups

227 separately. In addition to examining species of conservation concern and cup-nesters, we subset the

228 woodland assemblage data to remove the most dominant species (superb fairywren, yellow-rumped

229 thornbill, and willie wagtail).

230 For the woodland assemblage, and each subset, we followed a three-step modelling approach:

231 1. We first accounted for variation in our response variable associated with weather and

232 temporal factors. We incorporated variables of significance into subsequent models.

233 2. We then modelled our response variable against site type, comparing plantings, remnants,

234 and large reference sites.

235 3. Finally, we modelled our response variable against size and shape in plantings and

236 remnants, excluding large reference sites.

237 In each step, we fitted global models with all combinations of the variables of interest, and ranked

238 candidate models using Akaike’s Information Criterion corrected for small sample sizes (AICc). We

239 considered models with AICc ≤2 as top-ranked models (Burnham and Anderson, 2004). Weather

240 and temporal variables of significance identified in Step 1 were included in both Step 2 and Step 3.

12

241 We used the packages ‘lme4’ (Bates et al., 2015) and ‘MuMIn’ (Bartoń, 2018) in R version 3.4.4 (R

242 Core Team 2018) to fit and select models. Variograms were constructed using the package ‘geoR’

243 (Ribeiro and Diggle 2016).

244 Table 2 Linear mixed model parameters. The response variable is SCORE, and all 245 other variables are predictors.

Variable name Description

Square root of score of breeding activity recorded during surveys, calculated SCORE per Mac Nally (2007) and standardised by survey area (score/1.3 for small sites, score/3.0 for large sites) TYPE Site type (planting, remnant, reference) SIZE Site size (ha) SHAPE Measure of site shape, calculated as perimeter/width (m) AGE Age of planting at the commencement of the study (years) FENCED Site fenced from cattle (yes/no) Subjective measure of sun during surveys, on a numerical scale of 1-4 where 1 SUN = full sun and 4 = overcast Subjective measure of temperature during surveys, on a numerical scale of 1-8 TEMP where 1 = cold and 8 = hot Subjective measure of wind during surveys, on a numerical scale of 1-8 where WIND 1 = calm and 8 = strong wind TIME Time of day surveys commenced, given as no. hours post-sunrise (hr) DATE Julian date on which surveys were conducted 246

247 We used multivariate latent variable models from the package ‘boral’ (Hui, 2016) to compare how

248 abundance and breeding activity for bird species responded to site type, size, and shape. This

249 approach is useful because it allows for investigation of the association between multiple species

250 and underlying environmental variables in a linear modelling framework, while also accounting for

251 potential correlations among species. Specifically, we constructed one latent variable model for

252 each response matrix, and then compared the coefficient estimates for each species and variable.

253 For this modelling approach, only species detected both in point count surveys and breeding activity

254 surveys could be included. We subset our data to an assemblage of interest that included woodland-

255 dependent species (Silcocks et al. 2005) and several other small-bodied species that characterise the

256 bird community of woodlands in our study region (Appendix C). Due to the disproportionate spatial

257 influence of the frequently-detected superb fairywren in our initial ordination plots, we excluded it

258 from our multivariate latent variable models. 13

259 3. Results

260 3.1 General findings

261 A total of 90 bird species was detected during point count surveys, of which 66, or 73%, displayed

262 evidence of breeding activity (Appendix B). Additionally, two species – the hooded robin

263 (Melanodryas cucullata) and brown goshawk (Accipiter fasciatus) – were recorded in breeding

264 activity surveys but not detected in point counts. The most commonly detected species was the

265 superb fairywren, which accounted for 26% of all breeding activity recorded in the study. Other

266 frequently-detected species were the willie wagtail, yellow-rumped thornbill, grey shrikethrush

267 (Colluricincla harmonica), and rufous whistler (Pachycephala rufiventris). The species of

268 conservation concern we detected during surveys included the yellow-rumped thornbill, weebill

269 (Smicrornis brevirostris), speckled warbler (Pyrrholaemus sagittatus), dusky woodswallow

270 (Artamus cyanopterus), crested shrike-tit (Falcunculus frontatus), and hooded robin. For the

271 woodland assemblage, breeding activity scores recorded during surveys ranged from 11.5 to 104.5,

272 with a mean of 46.0 (n=105, SE=2.2). The mean score for cup-nesters was 19.1 (n=105, SE=1.4),

273 with minimum and maximum scores of 0 and 76.0 respectively. For species of conservation

274 concern, the mean score was 11.5 (n=105, SE=1.4), minimum score 0, and maximum score 55.0.

275 We found no differences in breeding activity in sites that were fenced compared with sites that were

276 exposed to grazing by stock.

277 3.2 How does woodland bird breeding activity in restoration plantings compare to breeding

278 activity in remnant woodland patches?

279 For the woodland assemblage, the score for breeding activity did not differ between plantings,

280 remnants, and reference sites (Appendix E). That is, site type did not appear as a variable of

281 significance in any of our top-ranked models. The same was true when comparing only plantings

282 and remnants (excluding reference sites) (Table 3). We found no effect of site type on species of

283 conservation concern, and cup-nesters showed no response to site type. Removing the superb

14

284 fairywren, willie wagtail and yellow-rumped thornbill from the woodland assemblage did not elicit

285 any response to site type from the remainder of the assemblage.

286 3.3 How do patch attributes affect breeding activity in plantings and remnant woodland patches?

287 Modelling patch attributes of remnants and plantings (excluding large reference sites) against

288 breeding activity score for the woodland assemblage revealed a strong negative effect of increasing

289 patch size, which appeared consistently in the top two candidate models (Table 3). That is, there

290 was more breeding activity per hectare in smaller patches (Figure 2). However, the removal of the

291 superb fairywren from the woodland assemblage greatly reduced the negative effect of site size on

292 breeding activity (Table 3). Size appeared as an explanatory variable in candidate models for

293 breeding activity score of assemblages without superb fairywren, willie wagtail, and yellow-rumped

294 thornbill, but its inclusion did not substantially improve the fit of the simplest model (containing

295 only Julian date). Where size appeared as an explanatory variable, its effect was marginal, with a

296 large standard error.

15

297 298 Figure 2 Effect plot illustrating the influence of patch size on breeding activity score of the woodland 299 assemblage in restoration plantings and similarly-sized woodland remnants. Shading indicates 95% 300 confidence intervals.

301 302 Excluding large reference sites revealed a marginal negative effect of site type, suggesting that

303 breeding activity score was higher in plantings than in similarly-sized woodland remnants (Table

304 3). However, the inclusion of site type did not substantially improve the fit of the simplest model,

305 and it failed to appear in top-ranked models after the removal of the three dominant species from

306 the assemblage. Shape appeared in one top-ranked model when the superb fairywren was excluded

307 from the assemblage, however, the standard error was larger than the effect size itself. Site shape

308 therefore had no interpretable effect on breeding activity score.

309 For species of conservation concern, the best fitting model was the model containing only Julian

310 date (Table 3). Consequently, there were no interpretable effects of planting size, shape, or type on

311 breeding activity score for this subset. The same result was observed for cup-nesters, again

16

312 indicating marginal or no effects of patch attributes on breeding activity score. Dome-nesters mirror

313 the negative response to patch size demonstrated by the woodland assemblage (per 0.81

314 correlation).

315 Table 3 Parameter estimates for total breeding score recorded during breeding activity surveys, ranked by 316 Akaike’s Information Criterion adjusted for small sample sizes (AICc). Top-ranked models (ΔAICc ≤2) are 317 shown for the woodland assemblage, species of conservation concern, cup-nesters, and subsets of the 318 woodland assemblage that exclude dominant species. All models that differed from the top model (ΔAICc) by 319 ≤2 are shown.

Woodland assemblage Rank 1 Rank 2 (w = 0.22) (w = 0.12)

Estimate (SE) Estimate (SE) Intercept 4.57 (0.22) 4.69 (0.24) DATE 0.59 (0.09) 0.58 (0.09) SIZE – 0.40 (0.16) – 0.39 (0.16) TYPE (remnant) – 0.37 (0.33)

Excluding superb Rank 1 Rank 2 Rank 3 fairywren (w = 0.19) (w = 0.11) (w = 0.08)

Estimate (SE) Estimate (SE) Estimate (SE) Intercept 3.83 (0.12) 3.83 (0.12) 3.69 (0.12) DATE 0.38 (0.10) 0.38 (0.10) 0.37 (0.10) SIZE – 0.13 (0.12) – 0.53 (0.21) SHAPE – 0.06 (0.11) SIZE:SHAPE – 0.67 (0.32)

Excluding superb Rank 1 Rank 2 Rank 3 fairywren, yellow-rumped (w = 0.20) (w = 0.09) (w = 0.09) thornbill, willie wagtail

Estimate (SE) Estimate (SE) Estimate (SE) Intercept 3.19 (0.15) 3.19 (0.14) 3.38 (0.27) DATE 0.24 (0.10) 0.24 (0.10) 0.24 (0.10) SIZE – 0.12 (0.14) FENCED (yes) – 0.26 (0.32)

Species of conservation Rank 1 Rank 2 Rank 3 Rank 4 Rank 5 Rank 6 concern (w = 0.13) (w = 0.11) (w = 0.07) (w = 0.06) (w = 0.06) (w = 0.05)

Estimate (SE) Estimate (SE) Estimate (SE) Estimate (SE) Estimate (SE) Estimate (SE) Intercept 1.76 (0.23) 1.99 (0.27) 1.38 (0.43) 1.76 (0.23) 1.99 (0.26) 2.00 (0.25) DATE 0.25 (0.10) 0.24 (0.10) 0.24 (0.10) 0.25 (0.10) 0.23 (0.10) 0.24 (0.10) TYPE (remnant) – 0.69 (0.47) – 0.71 (0.46) – 0.67 (0.43) FENCED (yes) 0.52 (0.51) SIZE 0.21 (0.23) 0.23 (0.22) 0.37 (0.23) SIZE:TYPE (remnant) – 0.77 (0.53)

Cup-nesters Rank 1 Rank 2 Rank 3 (w = 0.23) (w = 0.10) (w = 0.09)

Estimate (SE) Estimate (SE) Estimate (SE) Intercept 2.58 (0.21) 2.47 (0.25) 2.58 (0.20)

17

DATE 0.27 (0.10) 0.27 (0.10) 0.27 (0.10) TYPE (remnant) 0.32 (0.43) SHAPE 0.14 (0.21) 320

321 Planting age was a significant predictor of breeding activity for the woodland assemblage (Table 4).

322 An increase in planting age was associated with a decrease in breeding activity. This result was no

323 longer evident when the superb fairywren, yellow-rumped thornbill and willie wagtail were

324 removed from the assemblage. However, species of conservation concern also responded negatively

325 to an increase in planting age. No effect of planting age was observed for cup-nesters. For the latter

326 subset, the null model was the top-ranked model.

327 Table 4 Parameter estimates for total breeding score recorded during breeding activity surveys, ranked by 328 Akaike’s Information Criterion adjusted for small sample sizes (AICc). Top-ranked models (ΔAICc ≤2) are 329 shown for the woodland assemblage, species of conservation concern, and subsets of the woodland 330 assemblage that exclude dominant species. All models that differed from the top model (ΔAICc) by ≤2 are 331 shown. Note that candidate models for cup-nesters are not included, as the null model was the top-ranked 332 model for this subset.

Woodland assemblage Rank 1 Rank 2 Rank 3 (w = 0.27) (w = 0.22) (w = 0.10)

Estimate (SE) Estimate (SE) Estimate (SE) Intercept 5.65 (0.51) 4.73 (0.25) 4.73 (0.25) DATE 0.52 (0.11) 0.54 (0.11) 0.54 (0.11) AGE – 0.42 (0.15) – 0.51 (0.18) – 0.51 (0.18) SIZE – 0.37 (0.14) – 0.41 (0.18) – 0.45 (0.19) FENCED – 0.99 (0.53) SHAPE – 0.17 (0.19)

Excluding superb Rank 1 Rank 2 Rank 3 fairywren (w = 0.20) (w = 0.13) (w = 0.13)

Estimate (SE) Estimate (SE) Estimate (SE) Intercept 3.85 (0.12) 3.85 (0.14) 3.30 (0.42) DATE 0.32 (0.12) 0.31 (0.12) 0.32 (0.12) AGE – 0.25 (0.13) – 0.30 (0.12) FENCED 0.60 (0.45)

Excluding superb Rank 1 Rank 2 fairywren, yellow-rumped (w = 0.22) (w = 0.12) thornbill, willie wagtail

Estimate (SE) Estimate (SE) Intercept 3.16 (0.19) 3.16 (0.19) DATE 0.21 (0.11)

Species of conservation Rank 1 Rank 2 concern (w = 0.20) (w = 0.12)

Estimate (SE) Estimate (SE) 18

Intercept 1.99 (0.21) 1.99 (0.26) DATE 0.30 (0.13) 0.29 (0.13) AGE – 0.53 (0.21) – 0.52 (0.27) SIZE 0.43 (0.21) 333

334 3.4 Does breeding activity in restoration plantings and remnant woodland patches reflect species

335 assemblage composition?

336 Based on ordination modelling, we found that breeding activity was not strongly correlated with

337 relative abundance for bird species in our study sites (Figure 3). Examining the effects of patch

338 attributes on relative abundance and breeding activity revealed that many species differed in their

339 responses according to the two metrics. For example, the abundance of the willie wagtail in

340 remnants and plantings was similar, but breeding activity for this species was higher in remnants

341 (Figure 3a). A similar pattern was observed for the buff-rumped thornbill (Acanthiza reguloides),

342 for which more breeding activity was recorded in large reference sites than in plantings, despite the

343 species occurring in similar abundances in the two site types (Figure 3b). The buff-rumped thornbill

344 also displayed a positive response to increasing patch size according to relative abundance, but a

345 negative response according to breeding activity (Figure 3c). Interestingly, there were no species

346 whose abundance increased with patch linearity, but several species, including the black-faced

347 cuckooshrike (Coracina novaehollandiae), brown treecreeper (Climacteris picumnus), and willie

348 wagtail, showed more evidence of breeding activity in linear sites (Figure 3d). We note that

349 confidence intervals around the estimates for many species were large (Appendix H).

19

a) b)

350

c) d)

351

352 Figure 3 Bird species’ relative abundance and breeding activity plotted according to the effects of a) patch 353 type: remnants vs. plantings, in which a positive effect is associated with remnants, b) patch type: reference 354 sites vs. plantings, in which a positive effect is associated with reference sites, c) patch size, and d) patch 355 shape, in which the effect becomes more negative with increasing patch linearity. Effect sizes are taken from 356 multivariate latent variable models. ⚫ = species of conservation concern, ⚫ = species of least concern. 357

358 4. Discussion

359 We recorded breeding activity of a variety of bird species in both restoration plantings and remnant

360 woodland patches. Our analyses of the effects of patch attributes revealed several unexpected

361 findings – most notably, a negative effect of patch size driven by one dominant species (the superb

362 fairywren), in which there was more breeding activity per hectare in smaller patches. Inferences

20

363 from our study contrast with those of numerous studies on bird species richness and abundance in

364 fragmented agricultural landscapes, which report a positive effect of patch size. We further discuss

365 our key findings in the remainder of this paper and conclude with some insights for bird

366 conservation.

367 4.1 Patch size

368 Contrary to our predictions at the outset of this study based on patch size theory (Rosenzweig,

369 1995), we found that breeding activity score per hectare decreased as patch size increased in

370 plantings and remnant woodland patches. This result was driven by the most commonly-detected

371 species in the study region, and when this species was removed, there was no effect of patch size on

372 breeding activity. Both of these findings contrast with the majority of previous studies, which have

373 documented higher breeding success and reproductive output in larger habitat patches than in

374 smaller patches (e.g. Burke and Nol, 2000; Zanette, 2001; Zanette et al., 2000; Zanette and Jenkins,

375 2000). The value of small habitat patches for biodiversity in fragmented landscapes has been

376 highlighted via studies of bird species distribution and abundance (Fischer and Lindenmayer, 2002;

377 Gibbons and Boak, 2002; Le Roux et al., 2015; Manning et al., 2006), and was underscored by

378 Wintle et al. (2019) in their global synthesis of conservation studies. Our results indicate that small

379 patches may play a substantial role in supporting bird populations, which we discuss further in the

380 concluding sections of this paper.

381 According to island biogeography theory, which has been applied to fragmented agricultural

382 landscapes, smaller patches may stimulate a concentration effect of animal populations in

383 fragmented landscapes (MacArthur and Wilson, 1967). For example, waterbirds have been recorded

384 breeding in greater abundances on small versus large islands (Erwin et al., 1995). This may be

385 attributed to the relationship between resource distribution in the patch and surrounding matrix

386 (Estades, 2001). Animals may retreat from the poor quality matrix into habitat patches

387 (concentration effect), and then be reluctant to travel into the surrounding matrix (a so-called “fence

21

388 effect”). However, Connor et al. (2007) found that animal population densities tend to be positively

389 correlated with area, suggesting that density compensation may not be a common phenomenon.

390 Smaller patches, including plantings, have been found to contain bird communities with lower

391 overall species richness and a greater proportion of generalist or edge-specialist species (Flaspohler

392 et al., 2010; Mac Nally et al., 2010). Species that are tolerant to fragmentation may take advantage

393 of nesting habitat provided by small patches while utilising resources in the surrounding matrix

394 (Andrén, 1994; Driscoll et al., 2013; Estades, 2001). The superb fairywren accounted for over one

395 quarter of all observations of breeding activity in our study, and is often described as a habitat

396 generalist (Loyn et al., 2007; Mac Nally et al., 2010). Other commonly detected species, including

397 the willie wagtail, demonstrated a positive relationship between breeding activity and patch size,

398 indicating that not all species in the woodland assemblage respond similarly to patch size.

399 Furthermore, we found no effect of patch size on the collective group of species of conservation

400 concern.

401 Nest predation may have a significant influence on breeding success in birds, and can vary with

402 predator type, patch size, and isolation in fragmented landscapes (Okada et al., 2017; Stephens et

403 al., 2004). There is conflicting evidence pertaining to the influence of patch size on nest predation

404 in fragmented agricultural landscapes. For example, Hoover et al. (1995) attributed lower nesting

405 success of wood thrushes (Hylocichla mustelina) in smaller fragments to a greater abundance of

406 avian predators, and Major et al. (2001) found that the grey butcherbird (Cracticus torquatus), a

407 predatory species in Australian woodlands, was more abundant in smaller than in larger habitat

408 patches. In contrast, Zanette et al. (2000) found no evidence that area-sensitivity in the eastern

409 yellow robin (Eopsaltria australis) could be explained by nest predation. Lehnen and Rodewald

410 (2009) also found no evidence of area-sensitivity in survival and recruitment of shrubland bird

411 species of conservation concern in the eastern United States. Nest type is also confounded with

412 predation risk. Cup-nests are inherently more vulnerable to predation than dome-nests (Okada et al.,

22

413 2017), and thus species that build cup-nests may be more sensitive to edge-effects in smaller

414 patches. However, in our study, we found no evidence of a patch-size effect on cup-nesters.

415 Conversely, smaller patches may contain lower abundances of brood parasites such as Horsfield’s

416 bronzecuckoo (Chrysococcyx basalis) (Brooker and Brooker, 2003), reducing the risk of brood

417 parasitism. Indeed, cuckoos were detected infrequently in our study sites (Appendix B). Further

418 research is warranted to directly investigate nesting success of woodland birds in fragmented

419 agricultural landscapes.

420 A potential explanation for recording greater incidences of breeding activity in smaller patches than

421 in larger patches is that an observer may search smaller sites more thoroughly than larger ones

422 (Woolhouse, 1983). However, we used a search method standardised by area and time in an attempt

423 to control for potential effects of survey effort on activity detection rates. With an equivalent time

424 spent per unit area in each survey regardless of patch size, bias towards detecting more breeding

425 activity in smaller sites should not have influenced our results. However, we note that breeding

426 activity surveys are inherently biased towards species that nest in lower strata (such as the superb

427 fairywren).

428 4.2 Patch type

429 We predicted that remnant woodland patches would be characterised by more cases of successful

430 breeding than restoration plantings. However, our results showed that there was as much breeding

431 activity in restoration plantings as in remnant woodland patches. This result is somewhat

432 unexpected, as previous studies have found significant differences in bird species diversity and

433 abundance in plantings and remnants; remnants, and large remnants in particular, tend to support a

434 more diverse species assemblage than plantings (Arnold, 2003; Cunningham et al., 2008;

435 Lindenmayer et al., 2012; Loyn et al., 2007; Martin et al., 2011; Munro et al., 2011). Previous

436 studies of bird assemblages in fragmented agricultural landscapes have identified bird species that

437 are “planting specialists”, which preferentially occupy restoration plantings over remnant woodland

23

438 patches or other sites (reviewed by Belder et al., 2018). It was possible from the outset that breeding

439 activity in restoration plantings would be primarily accounted for by a select few of these species,

440 such as the generalist and edge-tolerant superb fairywren and willie wagtail. However, our

441 modelling indicated that the same trend may hold for species of conservation concern as well as the

442 woodland assemblage as a whole. This suggests that restoration plantings are providing habitat that

443 is as valuable for bird populations as remnant woodland patches. We note, however, that various

444 woodland-dependent species, including species of conservation concern such as the dusky

445 woodswallow and brown treecreeper, show a strong affinity for remnant woodland. We therefore

446 posit that restoration plantings play a complementary role in providing habitat for woodland birds,

447 and caution against restoration plantings being considered a direct replacement for remnant

448 woodland (see also Cunningham et al., 2007).

449 4.3 Patch shape

450 At this outset of this study, we predicted linear-shaped sites would support less breeding activity

451 than block-shaped sites. We found a weak negative association between patch linearity and bird

452 breeding activity in our study sites in only one candidate model, and therefore no strong evidence

453 that site shape influenced bird breeding activity in our study region. Previous studies have

454 suggested that increasing linearity negatively influences breeding birds (Helzer and Jelinski, 1999;

455 King et al., 2009). However, Selwood et al. (2009) found more evidence of successful breeding by

456 woodland birds in linear patches. Our ordination modelling revealed that some bird species (e.g.

457 brown treecreeper, black-faced cuckooshrike) showed more breeding activity in sites of increasing

458 linearity, even though this was not reflected in relative abundance. We suggest that further studies

459 are needed to confirm whether patch linearity influences breeding success of birds in fragmented

460 agricultural landscapes.

24

461 4.4 Planting age

462 Contrary to expectations, we found that planting age was a negative predictor of bird breeding

463 activity for the woodland assemblage and for species of conservation concern. That is, there was

464 less evidence of breeding activity in older plantings. This finding contrasts with that of Selwood et

465 al. (2009), who found that the age of plantings did not influence breeding activity. Barrett et al.

466 (2003) found evidence of bird breeding activity in plantings as young as three years, noting that the

467 species that exhibited the most breeding activity were small, shrub-swelling species such as the

468 superb fairywren, red-browed finch, and yellow-rumped thornbill.

469 A typical planting in our study region consists of a Eucalyptus overstorey and Acacia understorey.

470 In the absence of fire, an Acacia understorey is likely to senesce after 20-50 years (Broadhurst et

471 al., 2008; Parsons and Gosper, 2011), and natural regeneration of the shrub layer in planted sites

472 may be poor (Vesk et al., 2008). The deterioration of understorey density and diversity with

473 planting age is likely to contribute to a reduction in suitable nesting sites for common shrub-nesting

474 species like the superb fairywren, as well as species of conservation concern such as the yellow-

475 rumped thornbill and diamond firetail (Stagonopleura guttata). This may explain why the older

476 plantings in our study, which were around 25 years of age, did not support as much breeding

477 activity as younger plantings. The lack of an effect of age on the assemblage when our three most

478 dominant species were removed, as well as the absence of effects for cup-nesters, may be related to

479 the small sample size of these subsets.

480 4.5 Other findings

481 We found that examining breeding activity in our study sites provided a markedly different picture

482 of bird species’ responses to patch attributes than examining relative abundances obtained via point

483 counts. There were several species whose responses to patch size, shape, and type based on relative

484 abundance were opposite to their responses to these variables based on breeding activity. This

485 indicates that 1) some bird species choose to breed disproportionately more in particular kinds of

25

486 patches, or 2) the resources birds need to breed are not necessarily provided in patches that they

487 choose to forage in (Loyn et al., 2007). The latter is of particular interest, and important for

488 assessing the value of restoration plantings for woodland bird conservation; if birds preferentially

489 occupy habitat patches but are unable to breed successfully in them, then those patches may become

490 ecological traps, exacerbating population declines (Battin, 2004). This highlights the importance of

491 conducting research that moves beyond pattern-based data collection to include more detailed,

492 population-oriented studies (Belder et al., 2018; Ruiz-Jaen and Aide, 2005).

493 We found that for species of conservation concern, there were no interpretable effects of site type,

494 size, shape, or other variables on breeding activity score. The lack of an effect of site size is

495 surprising, as previous studies have found that site occupancy by species of conservation concern is

496 positively associated with patch size (Ford et al., 2009; Lindenmayer et al., 2010; Montague-Drake

497 et al., 2009).

498 The absence of any effect of site type was also unexpected, as we had predicted more breeding

499 activity by species of conservation concern in remnants due to the considerable body of evidence

500 indicating that many threatened and declining species are dependent on or closely associated with

501 remnant woodland (Cunningham et al., 2008; Kinross, 2004; Martin et al., 2011). We note that

502 some species of conservation concern, such as the yellow-rumped thornbill, are among “planting

503 specialists” identified in previous studies (Belder et al., 2018) (Appendix B). It is possible that the

504 small number of observations of species of conservation concern in our study reduced our power to

505 detect effects of patch attributes on these species, if they do indeed exist.

506 4.6 Inferential limitations

507 Variables at the landscape level, such as the amount and proximity of native vegetation, may have a

508 stronger influence on species richness and abundance (Cunningham et al., 2008; Fahrig, 2013;

509 Lindenmayer et al., 2010; Radford and Bennett, 2007) and breeding activity (Hinsley et al., 2008,

510 1995) than the patch-level characteristics of area and shape. Investigating these variables was

26

511 outside the scope of this study, but we recommend further research be undertaken to address their

512 effects. We note the prevalence of a select few species in our data, which may be symptomatic of an

513 environment that favours generalist and edge-tolerant species, to the detriment of richness and

514 productivity of woodland bird assemblages in our study region. Additionally, the absence of the

515 noisy miner in our study sites enabled us to examine the effects of patch attributes without the

516 confounding effects of competitive exclusion, but noisy miners are regular occupants of small

517 patches in fragmented agricultural landscapes (Major et al. 2001). We also note the small size (<10

518 ha) of plantings and remnants in our study. These reflect the typical size of native vegetation

519 patches in our study region, but we caution against applying our findings to much larger-scale

520 restoration projects, as breeding birds may respond differently to them than they do to small,

521 isolated patches. Lastly, we note the relatively short duration of our study – two breeding seasons in

522 years of above-average rainfall. We suggest that a better understanding of woodland bird population

523 processes could be obtained by incorporating breeding studies into long-term monitoring projects.

524 4.7 Management implications and concluding remarks

525 Studies of bird distribution and abundance in fragmented landscapes have previously highlighted

526 the conservation value of small habitat patches (Fischer and Lindenmayer, 2002; Flaspohler et al.,

527 2010; Gibbons and Boak, 2002; Wintle et al., 2019). Our results add credence to these findings by

528 providing evidence that birds not only occupy small patches, but display evidence of successful

529 breeding within them. Previous studies of bird species richness and abundance in restoration

530 plantings have recommended that plantings be as large as possible to maximise their conservation

531 value (Freudenberger, 2001; Watson et al., 2001; Westphal et al., 2007). We do not seek to

532 undermine the conservation value of very large-scale restoration projects, which were outside the

533 scope of this study, and we fully support the planting of large areas of native vegetation as a

534 strategy to increase vegetation cover in fragmented agricultural landscapes. However, our results

535 suggest that the establishment and conservation of small plantings (and the conservation of small

536 remnants) can also be of considerable value for the management of woodland bird populations (see 27

537 also Schippers et al., 2009). It is often easier and more cost-effective to implement and maintain

538 small patches (Kendal et al., 2017), so we are hopeful that our findings will encourage land

539 managers to consider implementing small plantings wherever it is not possible to establish large

540 plantings.

541 The observation of similar levels of breeding activity among the different site types in our study can

542 be cautiously interpreted as encouraging for the conservation value of restoration plantings, as it

543 indicates that birds in fragmented agricultural landscapes may view restoration plantings and

544 remnant woodland patches as equally suitable breeding habitat. However, we acknowledge that

545 breeding activity is only a proxy for breeding success, and cannot provide a true indication of

546 whether breeding attempts are succeeding or failing. We therefore recommend further exploration

547 using an approach such as monitoring nest success or daily nest survival.

548 Our finding that breeding activity decreased with planting age is of considerable interest for the

549 management of restoration plantings. If a reduction in the condition and density of the shrub layer

550 decreases the ability of a planting to support breeding birds, including species of conservation

551 concern, then there is a case for active management of the shrub layer (including replanting if

552 necessary) as a planting matures. Maintaining a complex habitat structure in restoration plantings

553 also decreases the likelihood of colonisation by the noisy miner (Kinross and Nicol, 2004; Maron et

554 al., 2013). Although we did not find evidence that fenced sites supported more breeding activity,

555 previous studies have shown that the ecological benefits of restoration plantings are diminished

556 when they are exposed to grazing by stock (Lindenmayer et al., 2018b; Selwood et al., 2009). We

557 suggest that maintaining fences around restoration plantings may assist with preserving the shrub

558 layer and ensuring that plantings continue to support breeding birds as they mature.

559 Finally, the unexpected nature of several of our key findings exemplifies the value of moving

560 beyond pattern data (such as site occupancy information) towards a more behaviour- and

561 population-oriented approach in monitoring and assessing the conservation value of restoration

28

562 plantings and other habitat patches. As we have shown, relying solely on measures like species

563 richness and abundance risks perpetuating critical knowledge gaps regarding habitat-use and the

564 value of habitat patches for birds in fragmented agricultural landscapes.

565 Acknowledgements

566 We are grateful for the feedback of three anonymous reviewers, which enabled us to make

567 substantial improvements to an earlier version of our manuscript. Ashwin Rudder and Ira Dudley-

568 Bestow assisted with fieldwork. We thank the landholders who permitted access to their properties

569 for the duration of the study. This work was funded by a Margaret Middleton Fund Award and

570 Lesslie Research Scholarship to DJB, with supporting funds from the Riverina Local Land Services.

571 Fieldwork was conducted under NSW Scientific Licence SL101588 and with approval from The

572 Australian National University’s Animal Ethics Committee.

573

574

29

575 Appendices

576 Appendix A Attributes of study sites in the South-west Slopes bioregion.

site site type shape planting elevation size (ha) perimeter width (m) fenced year (m) (m)

WB-6 planting block 2001 330 1.5 678 70 SU-5 planting block 2003 330 1.8 713 65  BL-4 planting block 1990 270 2.5 1583 70  HI-3 planting block 1990 285 5.3 983 150  MT-1 planting block 1991 280 5.6 988 200  RI-5 planting block 2002 249 7.7 1176 200  PM-4 planting linear 1997 269 1.3 1299 40  PS-3 planting linear 1989 429 1.4 1054 40  MH-6 planting linear 2000 344 1.7 1138 30  FR-A planting linear 1997 259 3.0 2610 15  FR-3 planting linear 1997 300 3.2 1655 40  MT-3 planting linear 1993 276 3.2 2213 30  SO-1 remnant block 303 2.1 1156 80 WS-3 remnant block 325 2.8 1006 130 PK-2 remnant block 265 5.8 1755 200  WB-2 remnant linear 262 2.3 1448 30 SU-1 remnant linear 297 4.1 736 60 PK-1 remnant linear 248 3.6 2379 25  GD-4 reference block 397 47.1 3956 555  KY reference block 347 110 5070 400  MG reference block 298 86 5090 400  577

30

578 Appendix B Assemblages and attributes of bird species recorded during the study. Breeding activity scores are provided for each species and site type. The 579 number of patches in which the species was detected and in which breeding occurred are provided in brackets: (no. patches breeding/no. patches present). • 580 denotes species recorded in point count surveys but not breeding activity surveys. Species are listed in taxonomic order (Gill and Donsker, 2018). Conservation 581 status according to NSW threatened species listing (NSW Environment & Heritage, 2018) and bird atlas trends (Barrett et al. 2003). Categories are least concern 582 (LC), conservation concern (CC), vulnerable (V). Information on breeding season and nest type taken from Morcombe (2003) and Pizzey and Knight (1997).

Species Abbreviation Nest type Breeding Conservation Plantings Remnants Reference sites season status

stubble quail Coturnix pectoralis SQ cup Aug-Mar LC • (0/2) • (0/1) wedge-tailed eagle Aquila audax WTE cup Jun-Nov LC • (0/1) brown goshawk Accipiter fasciatus BGOS cup Sep-Dec LC 3.0 (1/0) 1.0 (1/0) nankeen kestrel Falco cenchroides NK hollow Aug-Dec LC 1.0 (1/0) • (0/1) brown falcon Falco berigora BRFA cup Aug-Nov LC • (0/2) 7.0 (1/1) painted buttonquail Turnix varius PBQ cup Aug-Feb LC • (0/1) common bronzewing Phaps chalcoptera CBZ cup Aug-Dec LC 6.0 (1/3) 6.0 (1/2) crested pigeon Ocyphaps lophotes CP cup Jul-Dec LC 48.5 (3/3) 6.0 (2/5) peaceful dove Geopelia placida PD cup Oct-Jan LC 2.0 (1/4) 12.0 (2/4) gang-gang cockatoo Callocephalon fimbriatum GGC hollow Oct-Jan V 7.5 (1/1) galah Eolophus roseicapilla GAL hollow Jul-Dec LC • (0/11) • (0/6) • (0/3) little corella Cacatua sanguinea LCOR hollow Aug-Nov LC • (0/4) • (0/3) • (0/1) sulphur-crested cockatoo Cacatua galerita SCC hollow Aug-Jan LC • (0/6) • (0/4) • (0/3) crimson rosella Platycercus elegans CRO hollow Sep-Jan LC 8.5 (2/7) • (0/2) eastern rosella Platycercus eximius ERO hollow Aug-Dec LC 1.0 (1/12) 12.5 (2/6) • (0/2) red-rumped parrot Pseophotus haematonotus RRP hollow Aug-Jan LC 4.0 (2/8) • (0/5) • (0/1) Australian king-parrot Alisterus scapularis AKP hollow Sep-Jan LC • (0/1) 1.0 (1/1) • (0/1) superb parrot Polytelis swainsonii SUPA hollow Sep-Dec V • (0/2) • (0/2) Horsfield’s bronzecuckoo Chrysococcyx basalis HBC parasitic Aug-Jan LC • (0/1) • (0/2) shining bronzecuckoo Chrysococcyx lucidus SBC parasitic Aug-Jan LC • (0/1) • (0/1) pallid cuckoo Cacomantis pallidus PAC parasitic Aug-Dec LC • (0/1) fan-tailed cuckoo Cacomantis flabelliformis FTC parasitic Jul-Jan LC • (0/1) • (0/1) • (0/1) laughing kookaburra Dacelo novaeguineae LK hollow Sep-Dec LC • (0/11) 11.0 (2/4) 5.0 (1/1) sacred kingfisher Todiramphus sanctus SK hollow Sep-Jan LC • (0/2) 2.0 (2/3) 8.0 (1/2) rainbow bee-eater Merops ornatus RBE hollow Oct-Jan CC • (0/3) • (0/2) • (0/1) white-throated treecreeper Cormobates leucophaea WTTC hollow Aug-Jan LC • (0/1) • (0/1) 1.0 (1/3) brown treecreeper Climacteris picumnus BTC hollow May-Dec V 29.0 (2/4) 26.0 (1/3)

superb fairywrenP Malurus cyaneus SFW dome Sep-Dec LC 831.5 (12/12) 262.5 (6/12) 146.5 (3/12) little friarbird Philemon citreogularis LFB cup Jul-Nov LC 10.0 (1/2)

31 noisy friarbird Philemon corniculatus NFB cup Jul-Jan LC • (0/6) • (0/1) 9.0 (1/3) blue-faced honeyeater Entomyzon cyanotis BFH cup Jul-Jan LC • (0/2) black-chinned honeyeater Melithreptus gularis BCH cup Jul-Dec V 7.5 (1/1) • (0/2) brown-headed honeyeater Melithreptus brevirostris BHH cup Aug-Jan LC • (0/5) 7.5 (1/2) 5.0 (1/3) red wattlebirdP Anthochaera carunculata RWB cup Jul-Dec LC 15.0 (2/9) • (0/3) • (0/3) yellow-faced honeyeater Caligavis chrysops YFH cup Jul-Jan LC • (0/3) • (0/1) noisy miner Manorina melanocephala NM cup Jul-Dec LC 4.0 (2/6) 5.0 (2/3) fuscous honeyeater Ptilotula fusca FUH cup Aug-Dec LC 1.0 (0/1) white-plumed honeyeaterP Ptilotula penicillata WPH cup Aug-Dec LC 21.0 (3/11) 76.0 (3/6) • (0/3) spotted pardalote Pardalotus punctatus SPP hollow Sep-Dec LC 2.0 (1/3) • (0/1) striated pardalote Pardalotus striatus STP hollow Jun-Jan LC 11.0 (4/12) 1.0 (1/6) • (0/3) speckled warblerP Pyrrholaemus sagittatus SW dome Aug-Jan CC 1.0 (1/1) 2.0 (1/2) 51.5 (3/2) white-browed scrubwrenP Sericornis frontalis WBS dome Jul-Dec LC 25.0 (1/1) weebillP Smicrornis brevirostris WEE dome Aug-Feb CC 97.0 (7/9) 7.0 (2/2) 23.0 (2/2) western gerygoneP Gerygone fusca WEG dome Aug-Nov LC 26.5 (3/10) 2.0 (1/4) 36.5 (3/3) white-throated gerygone Gerygone olivacea WTG dome Sep-Nov LC • (0/3) 1.0 (1/2) brown thornbill Acanthiza pusilla BT dome Aug-Dec LC 2.0 (2/1) • (0/1) 54.0 (1/3) buff-rumped thornbill Acanthiza reguloides BRT dome Aug-Dec LC 32.5 (2/4) 55.5 (2/2) 155.5 (3/3) yellow-rumped thornbillP Acanthiza chrysorrhoa YRT dome Jul-Dec CC 513.5 (10/10) 53.0 (2/3) 48.5 (2/3) yellow thornbillP Acanthiza nana YET dome Aug-Dec LC 31.0 (8/9) 8.0 (2/2) 6.0 (2/3) striated thornbillP Acanthiza lineata STT dome Jul-Dec LC 17.5 (1/2) 13.0 (1/1) 31.5 (2/3) white-browed babbler Pomatostomus superciliosus WBB dome Jun-Dec LC 7.5 (1/2) 16.0 (1/1) grey butcherbird Cracticus torquatus GBB cup Aug-Dec LC • (0/2) • (0/2) pied butcherbird Cracticus nigrogularis PBB cup Aug-Nov LC 3.5 (1/10) 7.5 (1/5) • (0/1) Australian magpie Cracticus tibicen AM cup Aug-Oct LC 143.5 (8/12) 74.5 (4/6) 19.5 (2/3) pied currawong Strepera graculina PCW cup Aug-Dec LC • (0/2) • (0/2) • (0/1) dusky woodswallow Artamus cyanopterus DWS cup Aug-Dec V 1.0 (1/0) 36.0 (2/3) 6.0 (1/1) black-faced cuckooshrike Coracina novaehollandiae BFCS cup Aug-Jan LC 28.5 (4/12) 32.0 (4/5) 2.0 (2/3) white-bellied cuckooshrike Coracina papuensis WBCS cup Aug-Mar LC • (0/1) white-winged triller Lalage tricolor WWT cup Sep-Dec CC 25.0 (2/1) 7.0 (1/2) 1.0 (1/2) varied sittella Daphoenositta chrysoptera VS cup Sep-Dec V 27.0 (3/3) crested shriketit Falcunculus frontatus CST cup Sep-Jan CC 5.0 (3/5) 9.0 (1/4) golden whistler Pachycephala pectoralis GOW cup Aug-Jan LC 23.5 (3/2) 2.0 (2/2) 7.5 (1/0) rufous whistlerP Pachycephala rufiventris RUW cup Sep-Feb LC 164.5 (7/12) 16.5 (3/5) 20.0 (3/3) grey shrikethrushP Colluricincla harmonica GST cup Jul-Feb LC 135.0 (9/12) 52.5 (4/6) 8.5 (2/3) olive-backed oriole Oriolus sagittatus OBO cup Sep-Jan LC 7.5 (1/0) • (0/1) willie wagtailP Rhipidura leucophrys WW cup Aug-Dec LC 92.5 (9/12) 176.5 (5/6) 42.5 (1/3) grey fantailP Rhipidura albiscapa GF cup Aug-Dec LC 60.5 (9/12) 24.0 (3/5) 17.0 (3/3) 32

magpie-lark Grallina cyanoleuca AML cup Aug-Feb LC 10.5 (4/12) 43.0 (2/6) 5.0 (1/3) leaden flycatcher Myiagra rubecula LFC cup Sep-Nov LC • (0/1) • (0/1) restless flycatcher Myiagra inquieta RFC cup Aug-Jan CC • (0/2) • (0/5) • (0/1) little raven Corvus mellori LR cup Aug-Dec LC 16.5 (2/8) 7.5 (1/4) 9.0 (1/2) Australian raven Corvus coronoides AR cup Jul-Oct LC 11.5 (1/12) 3.5 (1/6) 11.5 (2/3) white-winged chough Corcorax melanoramphos WWC cup Aug-Dec LC 55.5 (6/6) 40.0 (2/3) 15.5 (2/3)

eastern yellow robinP Eopsaltria australis EYR cup Jul-Dec LC • (0/1) 7.0 (1/1)

hooded robinP Melanodryas cucullata HR cup Jul-Dec V 4.5 (1) jacky winter Microeca fascinans JW cup Jul-Dec CC • (0/1) 10.0 (2/2) 37.0 (2/2)

flame robinP Petroica phoenicea FR cup Aug-Jan V • (0/1) • (0/1)

red-capped robinP Petroica goodenovii RCR cup Jul-Jan CC 95.5 (2/2) • (0/1) 36.0 (3/1) welcome swallow Hirundo neoxena WS cup Aug-Dec LC • (0/4) • (0/6) • (0/1) fairy martin Petrochelidon ariel FM other Aug-Jan CC • (0/1) tree martin Petrochelidon nigricans TM hollow Aug-Dec LC 8.5 (1/2) rufous songlark Cincloramphus mathewsi RSL cup Sep-Dec LC 6.0 (1/8) 16.0 (3/3) • (0/1) brown songlark Cincloramphus cruralis BSL cup Sep-Feb CC • (0/3) • (0/1) silvereye Zosterops lateralis SIL cup Sep-Jan LC 1.0 (1/5) • (0/2) • (0/1)

common starlingI Sturnus vulgaris STA hollow Aug-Jan 24.0 (2/8) 10.0 (1/5) 6.0 (1/1)

common blackbirdI Turdus merula BKB cup Sep-Dec 42.0 (2/3) • (0/1) mistletoebird Dicaeum hirundinaceum MTB dome Oct-Mar LC • (0/1) • (0/2) 1.0 (1/2)

diamond firetailP Stagonopleura guttata DF dome Aug-Jan V 38.0 (2/2) 10.0 (3/2)

red-browed finchP Neochmia temporalis RBF dome Sep-Dec LC 29.0 (2/3) 6.0 (1/1) double-barred finch Taeniopygia bichenovii DBF dome Jul-Dec LC 2.0 (2/2) 1.0 (1/1) Australian pipit Anthus australis PIP cup Aug-Dec LC • (0/3) • (0/2)

583 P Planting specialists (Belder et al. 2018)

584 I Introduced species

33

585 Appendix C Total breeding activity recorded for the subset of bird species included in multivariate latent 586 model ordinations. Acronyms corresponding to particular bird species are given in Appendix B.

1400

1200

1000

800

600

400 Total breeding activity score activity breeding Total

200

0

BT

DF VS

PD GF CP

JW

SIL

SW

LFB

WW STT

STP YET

RSL

DBF YRT BRT BTC RBF CST CBZ NFB SPP

GST EYR FUH

BCH BHH MTB

RCR

SFW

OBO

WEE WBS WBB

WPH DWS RWB WTG

RUW

WEG

GOW WWT

WWC BFCS 587 WTTC

588 589

34

590 Appendix D Mixed effects models for breeding score modelled against weather and temporal variables, 591 ranked by Akaike’s Information Criterion adjusted for small sample sizes (AICc). Top-ranked models (ΔAICc 592 ≤2) are shown for the woodland assemblage, species of conservation concern, cup-nesters, and subsets of 593 the assemblage that exclude dominant species. All models that differed from the top model (ΔAICc) by ≤2 594 are shown, as well as the intercept-only model.

Woodland assemblage df log(L) AICc ΔAICc AICw

DATE 5 -144.46 299.52 0.00 0.29 DATE + TIME 6 -143.57 300.00 0.49 0.23 DATE + SUN 6 -144.29 301.44 1.92 0.11 Intercept only 4 -160.21 328.82 29.30 0.00

Excluding superb fairywren df log(L) AICc ΔAICc AICw

DATE 5 -144.32 299.25 0.00 0.43 Intercept only 4 -150.68 309.75 10.50 0.00

Excluding superb fairywren, yellow-rumped thornbill, willie wagtail df log(L) AICc ΔAICc AICw

DATE 5 -145.52 301.64 0.00 0.29 DATE + TIME 6 -145.15 303.17 1.53 0.14 DATE + WIND 6 -145.30 303.46 1.82 0.12 Intercept only 4 -148.41 305.23 3.59 0.05

Species of conservation concern df log(L) AICc ΔAICc AICw

DATE + TIME 6 -157.23 327.32 0.00 0.31 DATE 5 -158.89 328.39 1.06 0.18 Intercept only 4 -161.53 331.45 4.13 0.04

Cup-nesters df log(L) AICc ΔAICc AICw

DATE + TIME 6 -150.37 313.60 0.00 0.23 DATE + TIME + WIND 7 -149.54 314.23 0.63 0.17 DATE + TIME + SUN 7 -149.66 314.48 0.88 0.15 DATE 5 -151.94 314.49 0.89 0.15 DATE + TIME + SUN + WIND 8 -148.96 315.42 1.82 0.09 DATE + SUN 6 -151.34 315.53 1.93 0.09 Intercept only 4 -155.94 320.29 6.69 0.01 595 596

35

597 Appendix E Mixed effects models for total breeding score modelled against site type for all sites (planting, 598 remnant, and reference), ranked by Akaike’s Information Criterion adjusted for small sample sizes (AICc). 599 Top-ranked models (ΔAICc ≤2) are shown for the woodland assemblage, species of conservation concern, 600 cup-nesters, and subsets of the woodland assemblage that exclude dominant species. All models that 601 differed from the top model (ΔAICc) by ≤2 are shown, as well as the intercept-only model.

Woodland assemblage df log(L) AICc ΔAICc AICw

DATE 5 -144.46 299.52 0.00 0.54 DATE + FENCED 6 -144.27 301.40 1.89 0.21 Intercept only 4 -160.21 328.82 29.30 0.00

Excluding superb fairywren df log(L) AICc ΔAICc AICw

DATE 5 -144.32 299.25 0.00 0.67 Intercept only 4 -150.68 309.75 10.50 0.00

Excluding superb fairywren, yellow-rumped thornbill, willie wagtail df log(L) AICc ΔAICc AICw

DATE 5 -145.52 301.64 0.00 0.47 Intercept only 4 -148.41 305.23 3.59 0.08

Species of conservation concern df log(L) AICc ΔAICc AICw

DATE 5 -158.89 328.39 0.00 0.34 DATE + FENCED 6 -158.10 329.05 0.67 0.24 DATE + TYPE 7 -157.23 329.62 1.24 0.18 Intercept only 4 -161.53 331.45 3.07 0.07

Cup-nesters df log(L) AICc ΔAICc AICw

DATE 5 -151.94 314.49 0.00 0.60 Intercept only 4 -155.94 320.29 5.79 0.03 602 603

36

604 Appendix F Mixed effects models for total breeding score recorded during breeding activity surveys in 605 plantings and remnants (excluding reference sites), ranked by Akaike’s Information Criterion adjusted for 606 small sample sizes (AICc). Top-ranked models (ΔAICc ≤2) are shown for the woodland assemblage, species 607 of conservation concern, cup-nesters, and subsets of the woodland assemblage that exclude dominant 608 species. All models that differed from the top model (ΔAICc) by ≤2 are shown, as well as the intercept-only 609 model.

Woodland assemblage df log(L) AICc ΔAICc AICw

DATE + SIZE 6 -122.29 257.58 0.00 0.22 DATE + SIZE + TYPE 7 -121.68 258.73 1.15 0.12 Intercept only 4 -140.70 289.88 32.29 0.00

Excluding superb fairywren df log(L) AICc ΔAICc AICw

DATE 5 -123.81 258.33 0.00 0.19 DATE + SIZE 6 -123.18 259.37 1.04 0.11 DATE + SIZE + SHAPE + SIZE:SHAPE 8 -121.17 260.12 1.79 0.08 Intercept only 4 -130.79 270.05 11.72 0.00

Excluding superb fairywren, yellow-rumped thornbill, willie wagtail df log(L) AICc ΔAICc AICw

DATE 5 -126.14 263.00 0.00 0.20 DATE + SIZE 6 -125.82 264.65 1.65 0.09 DATE + FENCED 6 -125.82 264.65 1.65 0.09 Intercept only 4 -129.16 266.79 3.79 0.03

Species of conservation concern df log(L) AICc ΔAICc AICw

DATE 5 -140.41 291.53 0.00 0.13 DATE + TYPE 6 -139.40 291.81 0.28 0.11 DATE + FENCED 6 -139.90 292.81 1.28 0.07 DATE + SIZE 6 -139.99 292.99 1.46 0.06 DATE + SIZE + TYPE 7 -138.84 293.05 1.52 0.06 DATE + SIZE + TYPE + SIZE:TYPE 8 -137.85 293.49 1.96 0.05 Intercept only 4 -143.08 294.63 3.10 0.03

Cup-nesters df log(L) AICc ΔAICc AICw

DATE 5 -133.19 277.10 0.00 0.23 DATE + TYPE 6 -132.92 278.85 1.75 0.10 DATE + SHAPE 6 -132.96 278.93 1.83 0.09 Intercept only 4 -136.61 281.69 4.59 0.02 610 611

37

612 Appendix G Mixed effects models for total breeding score recorded during breeding activity surveys in 613 plantings, ranked by Akaike’s Information Criterion adjusted for small sample sizes (AICc). Top-ranked 614 models (ΔAICc ≤2) are shown for the woodland assemblage, species of conservation concern, cup-nesters, 615 and subsets of the woodland assemblage that exclude dominant species. All models that differed from the 616 top model (ΔAICc) by ≤2 are shown, as well as the intercept-only model.

Woodland assemblage df log(L) AICc ΔAICc AICw

DATE + AGE + SIZE + FENCED 8 -76.62 172.06 0.00 0.27 DATE + AGE + SIZE 7 -78.16 172.47 0.40 0.22 DATE + AGE + SIZE + SHAPE 8 -77.60 174.01 1.95 0.10 Intercept only 4 -92.38 193.49 21.43 0.00

Excluding superb fairywren Df log(L) AICc ΔAICc AICw

DATE + AGE 6 -79.79 173.16 0.00 0.20 DATE 5 -81.45 174.00 0.84 0.13 DATE + AGE + FENCED 7 -78.95 174.06 0.90 0.13 Intercept only 4 -84.72 178.18 5.01 0.02

Excluding superb fairywren, yellow-rumped thornbill, willie wagtail Df log(L) AICc ΔAICc AICw

DATE 5 -82.35 175.81 0.00 0.22 Intercept only 4 -84.12 176.96 1.15 0.12

Species of conservation concern df log(L) AICc ΔAICc AICw

DATE + AGE + SIZE 7 -90.30 196.75 0.00 0.20 DATE + AGE 6 -92.08 197.75 1.00 0.12 Intercept only 4 -96.31 201.35 4.59 0.02

Cup-nesters df log(L) AICc ΔAICc AICw

Intercept only 4 -91.13 190.98 0.00 0.17 AGE 5 -90.42 191.94 0.96 0.10 DATE 5 -90.48 192.06 1.08 0.10 617

38

618 Appendix H Coefficients from multivariate latent variable models used to plot the effects of site attributes on relative abundance and breeding activity of bird 619 species. The 95% upper and lower confidence limits around the estimate are provided in brackets. Estimates for which the confidence interval does not overlap zero 620 are shown in bold.

ABUNDANCE BREEDING ACTIVITY

remnant vs. reference vs. remnant vs. reference vs. common name patch size patch linearity patch size patch linearity planting planting planting planting

black-faced cuckooshrike 0.17 (-0.36, 0.80) 1.11 (-0.30, 2.61) -0.21 (-0.76, 0.34) -0.04 (-0.34, 0.21) 0.09 (-0.02, 0.21) 0.00 (-0.39, 0.32) 0.01 (-0.13, 0.14) 0.05 (0.00, 0.09) brown-headed honeyeater -0.67 (-1.86, 0.46) -0.87 (-3.62, 1.59) 0.41 (-0.46, 1.42) -1.61 (-3.13, -0.59) 0.00 (-0.04, 0.04) 0.19 (0.06, 0.29) -0.06 (-0.10, -0.01) 0.00 (-0.01, 0.01) buff-rumped thornbill -0.19 (-1.25, 0.98) -0.66 (-2.64, 1.36) 1.28 (0.58, 2.03) -1.49 (-2.86, -0.22) 0.14 (-0.08, 0.36) 1.11 (0.36, 1.76) -0.11 (-0.37, 0.18) -0.08 (-0.18, 0.00) brown thornbill 0.08 (-2.13, 1.96) 3.63 (0.82, 6.17) -0.23 (-1.14, 0.71) 0.10 (-1.04, 1.15) 0.00 (-0.13, 0.11) 1.67 (1.26, 2.13) -0.50 (-0.65, -0.34) -0.01 (-0.05, 0.05)

brown treecreeper 4.57 (2.56, 6.97) 0.60 (-2.67, 4.10) 1.50 (0.65, 2.42) 0.09 (-0.32, 0.51) 0.10 (0.04, 0.15) -0.01 (-0.19, 0.18) 0.02 (-0.04, 0.09) 0.03 (0.01, 0.05) crested pigeon 1.58 (0.53, 2.72) -0.56 (-6.02, 5.19) -0.94 (-4.46, 1.42) 0.20 (-0.26, 0.68) 0.00 (-0.07, 0.07) -0.03 (-0.23, 0.20) 0.00 (-0.08, 0.07) 0.01 (-0.02, 0.03) double-barred finch -0.29 (-1.94, 1.45) -2.15 (-7.56, 2.84) -0.50 (-3.18, 1.57) -1.48 (-3.00, -0.35) 0.00 (-0.01, 0.01) -0.01 (-0.05, 0.04) 0.00 (-0.01, 0.02) 0.00 (-0.01, 0.00)

diamond firetail -0.71 (-3.20, 0.78) -0.89 (-6.45, 3.72) -1.35 (-4.36, 1.22) -5.67 (-7.29, -2.35) -0.05 (-0.15, 0.04) -0.07 (-0.32, 0.23) -0.01 (-0.11, 0.09) -0.03 (-0.06, 0.00) dusky woodswallow 4.79 (2.91, 6.69) -2.01 (-6.44, 2.55) 1.59 (-0.11, 3.24) -0.18 (-0.93, 0.69) 0.10 (0.04, 0.16) -0.05 (-0.23, 0.17) 0.02 (-0.07, 0.08) -0.02 (-0.04, 0.01) grey fantail -1.01 (-1.56, -0.48) -0.05 (-1.29, 1.03) 0.14 (-0.27, 0.58) -0.46 (-0.76, -0.18) 0.08 (-0.01, 0.18) 0.11 (-0.19, 0.41) -0.02 (-0.13, 0.09) -0.05 (-0.09, -0.01) golden whistler -0.47 (-2.93, 1.57) -1.31 (-6.81, 3.88) -1.98 (-4.89, 0.68) -1.19 (-2.73, 0.41) -0.03 (-0.15, 0.06) -0.27 (-0.59, 0.04) 0.12 (0.00, 0.13) -0.02 (-0.06, 0.02)

grey shrikethrush -0.29 (-0.66, 0.13) 0.67 (-0.59, 1.96) -0.44 (-0.94, 0.06) -0.07 (-0.26, 0.13) -0.04 (-0.21, 0.11) -0.01 (-0.15, 0.19) -0.04 (-0.23, 0.13) 0.00 (-0.06, 0.06)

jacky winter 2.22 (0.93, 3.76) 2.38 (-0.16, 4.64) 0.10 (-0.71, 0.89) -3.64 (-5.06, -2.38) 0.06 (-0.04, 0.14) -0.08 (-0.38, 0.21) 0.11 (0.01, 0.23) -0.01 (-0.05, 0.02) peaceful dove 1.75 (0.92, 2.68) -0.68 (-5.18, 3.70) -1.39 (-4.14, 0.59) -0.21 (-0.71, 0.33) 0.07 (0.01, 0.13) 0.01 (-0.15, 0.19) 0.00 (-0.07, 0.07) 0.02 (-0.01, 0.04) red-capped robin -2.71 (-4.69, -0.39) -3.03 (-7.72, 0.50) 0.59 (-1.33, 2.34) -3.71 (-5.39, -2.48) -0.18 (-0.34, -0.02) -0.49 (-1.05, -0.04) 0.09 (-0.09, 0.30) -0.09 (-0.16, -0.03)

rufous whistler -0.89 (-1.30, -0.48) 0.63 (-0.38, 1.64) -0.09 (-0.45, 0.29) -0.23 (-0.43, -0.02) -0.16 (-0.33, -0.03) -0.13 (-0.56, 0.35) 0.01 (-0.17, 0.17) 0.03 (-0.02, 0.10) red wattlebird -0.49 (-1.06, 0.13) -3.48 (-6.60, -1.03) 1.12 (0.31, 2.01) 0.01 (-0.24, 0.30) -0.02 (-0.06, 0.02) -0.01 (-0.15, 0.11) 0.00 (-0.05, 0.05) 0.01 (-0.01, 0.03)

sacred kingfisher 1.78 (0.84, 2.85) -1.34 (-4.55, 1.52) 1.17 (0.25, 2.17) -0.90 (-1.82, -0.03) 0.01 (-0.05, 0.06) -0.17 (-0.37, 0.00) 0.10 (0.03, 0.17) 0.00 (-0.02, 0.03)

striated pardalote 0.11 (-0.33, 0.62) -2.27 (-4.85, 0.14) 0.46 (-0.28, 1.48) 0.13 (-0.06, 0.35) -0.01 (-0.04, 0.01) -0.01 (-0.09, 0.07) 0.00 (-0.03, 0.03) 0.00 (-0.01, 0.01) striated thornbill -0.61 (-2.61, 1.19) 3.52 (1.05, 6.37) -0.12 (-1.24, 0.81) -2.06 (-4.95, -0.29) 0.03 (-0.12, 0.17) -0.02 (-0.46, 0.37) 0.09 (-0.06, 0.24) -0.02 (-0.07, 0.03)

speckled warbler -1.18 (-3.06, 0.41) 1.19 (-1.81, 4.39) -0.11 (-1.32, 1.13) -2.69 (-4.29, -1.07) 0.00 (-0.13, 0.13) 0.06 (-0.30, 0.48) 0.14 (-0.01, 0.28) 0.00 (-0.05, 0.05) varied sittella -1.65 (-4.50, 1.29) 3.10 (-0.10, 6.34) 0.14 (-0.84, 1.23) -1.54 (-4.17, 0.57) 0.00 (-0.08, 0.08) 0.06 (-0.16, 0.30) 0.05 (-0.03, 0.13) 0.00 (-0.03, 0.03)

white-browed scrubwren -3.00 (-7.08, 0.28) -2.34 (-7.36, 3.75) -4.09 (-8.19, -0.80) -0.77 (-3.26, 1.14) -0.05 (-0.14, 0.04) -0.05 (-0.30, 0.27) -0.01 (-0.11, 0.13) -0.01 (-0.05, 0.02)

39

weebill -0.97 (-1.57, -0.33) -5.07 (-6.81, -3.25) 1.90 (1.24, 2.47) -0.67 (-1.01, -0.38) -0.13 (-0.27, 0.01) -0.32 (-0.80, 0.12) 0.08 (-0.08, 0.26) -0.04 (-0.09, 0.02) western gerygone -0.91 (-1.62, -0.18) 0.16 (-1.39, 1.56) -0.01 (-0.60, 0.52) -0.88 (-1.30, -0.50) -0.06 (-0.17, 0.05) -0.11 (-0.45, 0.25) 0.11 (-0.02, 0.24) -0.02 (-0.06, 0.02)

white-plumed honeyeater 0.77 (0.34, 1.12) -4.75 (-7.35, -2.30) 1.51 (0.77, 2.36) 0.09 (-0.10, 0.31) 0.14 (0.03, 0.24) -0.01 (-0.34, 0.28) 0.00 (-0.11, 0.13) 0.02 (-0.02, 0.06) willie wagtail 0.53 (0.15, 0.91) -0.49 (-1.95, 1.13) 0.07 (-0.43, 0.63) 0.04 (-0.15, 0.24) 0.47 (0.29, 0.66) -0.42 (-1.03, 0.17) 0.22 (0.02, 0.45) 0.03 (-0.05, 0.02) white-winged chough -0.52 (-1.40, 0.55) -0.73 (-3.26, 1.75) 0.46 (-0.36, 1.36) -1.34 (-2.19, -0.62) 0.03 (-0.07, 0.13) -0.01 (-0.31, 0.28) 0.04 (-0.06, 0.16) -0.01 (-0.05, 0.02) yellow thornbill -1.39 (-2.05, -0.74) 0.03 (-1.55, 1.66) -0.30 (-0.98, 0.34) -0.39 (-0.68, -0.13) -0.04 (-0.11, 0.05) -0.15 (-0.38, 0.09) 0.07 (-0.02, 0.15) 0.00 (-0.03, 0.03) yellow-rumped thornbill -1.66 (-2.29, -1.08) 0.07 (-1.53, 1.49) -0.44 (-1.07, 0.19) -0.19 (-0.42, 0.04) -0.54 (-0.84, -0.24) -0.59 (-1.59, 0.38) 0.00 (-0.39, 0.34) -0.07 (-0.19, 0.04) 621

40

622 References

623 Andrén, H., 1994. Effects of Habitat Fragmentation on Birds and Mammals in Landscapes with 624 Different Proportions of Suitable Habitat: A Review. Oikos 71, 355. 625 https://doi.org/10.2307/3545823 626 Arnold, G.W., 2003. Bird species richness and abundance in wandoo woodland and in tree 627 plantations on farmland at Baker’s Hill, Western Australia. Emu 103, 259–269. 628 https://doi.org/http://dx.doi.org/10.1071/MU01005 629 Barral, M.P., Rey Benayas, J.M., Meli, P., Maceira, N.O., 2015. Quantifying the impacts of 630 ecological restoration on biodiversity and ecosystem services in agroecosystems: A global 631 meta-analysis. Agric. Ecosyst. Environ. 202, 223–231. 632 https://doi.org/http://dx.doi.org/10.1016/j.agee.2015.01.009 633 Barrett, G., Silcocks, A., Barry, S., Cunningham, R., and Poulter, R., 2003. The new atlas of 634 Australian birds. Royal Australasian Ornithologists Union, Hawthorn East, Victoria. 635 Barrett, G.W., Freudenberger, D., Drew, A., Stol, J., Nicholls, A.O., Cawsey, E.M., 2008. 636 Colonisation of native tree and shrub plantings by woodland birds in an agricultural landscape. 637 Wildl. Res. 35, 19–32. https://doi.org/Doi 10.1071/Wr07100 638 Bartoń, K., 2018. MuMIn: Multi-Model Inference. R package version 1.40.4. https://cran.r- 639 project.org/package=MuMIn 640 Bates, D., Maechler, M., Bolker, B., Walker, S., 2015. Fitting Linear Mixed-Effects Models Using 641 lme4. J. Stat. Softw. 67, 1–48. 642 Battin, J., 2004. When good animals love bad habitats: Ecological traps and the conservation of 643 animal populations. Conserv. Biol. 18, 1482–1491. https://doi.org/10.1111/j.1523- 644 1739.2004.00417.x 645 Belder, D.J., Pierson, J.C., Ikin, K., Lindenmayer, D.B., 2018. Beyond pattern to process: current 646 themes and future directions for the conservation of woodland birds through restoration 647 plantings. Wildl Res 45:473. doi: 10.1071/wr17156 648 Bennett, J.M., Clarke, R.H., Thomson, J.R., Mac Nally, R., 2015. Fragmentation, vegetation change 649 and irruptive competitors affect recruitment of woodland birds. Ecography (Cop.). 38, 163– 650 171. https://doi.org/10.1111/ecog.00936 651 Broadhurst, L.M., Young, A.G., Forrester, R., 2008. Genetic and demographic responses of 652 fragmented Acacia dealbata (Mimosaceae) populations in southeastern Australia. Biol Conserv 653 141:2843–2856. https://doi.org/10.1016/J.BIOCON.2008.08.021 654 Brooker, M., Brooker, L., 2003. Brood parasitism by Horsfield’s Bronze-Cuckoo in a fragmented 655 agricultural landscape in Western Australia. Emu 103, 357–361. 656 https://doi.org/http://dx.doi.org/10.1071/MU02034 657 Burke, D.M., Nol, E., 2000. Landscape and fragment size effects on reproductive success of forest- 658 breeding birds in Ontario. Ecol. Appl. 10, 1749–1761. https://doi.org/10.1890/1051- 659 0761(2000)010[1749:LAFSEO]2.0.CO;2 660 Burnham, K.P., Anderson, D.R., 2004. Multimodel inference: understanding AIC and BIC in model 661 selection. Sociol. Methods Res. 33, 261–304. https://doi.org/10.1177/0049124104268644 662 Chalfoun, A.D., Martin, T.E., 2007. Assessments of habitat preferences and quality depend on 663 spatial scale and metrics of fitness. J. Appl. Ecol. 44, 983–992. 664 Connor, E.F., Courtney, A.C., Yoder, J.M., 2007. Individuals-Area Relationships: The Relationship 665 between Animal Population Density and Area. Ecology 81, 734–748. 666 https://doi.org/10.1890/0012-9658(2000)081[0734:IARTRB]2.0.CO;2 667 Crouzeilles, R., Curran, M., Ferreira, M.S., Lindenmayer, D.B., Grelle, C.E. V., Rey Benayas, J.M., 668 2016. A global meta-analysis on the ecological drivers of forest restoration success. Nat. 669 Commun. 7, 11666. https://doi.org/10.1038/ncomms11666 670 Cunningham, R.B., Lindenmayer, D.B., Crane, M.J., Michael, D.R., Barton, P.S., Gibbons, P., 41

671 Okada, S., Ikin, K., Stein, J.A.R., 2014. The law of diminishing returns: woodland birds 672 respond to native vegetation cover at multiple spatial scales and over time. Divers. Distrib. 20, 673 59–71. https://doi.org/10.1111/ddi.12145 674 Cunningham, R.B., Lindenmayer, D.B., Crane, M.J., Michael, D.R., MacGregor, C.I., 2007. Reptile 675 and arboreal marsupial response to replanted vegetation in agricultural landscapes. Ecol. Appl. 676 17, 609–619. https://doi.org/Doi 10.1890/05-1892 677 Cunningham, R.B., Lindenmayer, D.B., Crane, M.J., Michael, D.R., MacGregor, C.I., Montague- 678 Drake, R.M., Fischer, J., 2008. The combined effects of remnant vegetation and tree planting 679 on farmland birds. Conserv. Biol. 22, 742–752. 680 Debus, S.J.S., Martin, W.K., Lemon, J.M., 2017. Changes in woodland bird communities as 681 replanted woodland matures. Pacific Conserv Biol 23:359–371. 682 https://doi.org/10.1071/PC16028 683 Department of the Environment, 2018. White Box-Yellow Box-Blakely’s Red Gum Grassy 684 Woodland and Derived Native Grassland in Community and Species Profile and Threats 685 Database [WWW Document]. Dep. Environ. Energy, Canberra. URL 686 http://www.environment.gov.au/sprat (accessed 8.11.18). 687 Dray, S., Dufour, A.-B., 2007. The ade4 package: Implementing the duality diagram for ecologists. 688 J. Stat. Softw. 22, 1–20. 689 Driscoll, D.A., Banks, S.C., Barton, P.S., Lindenmayer, D.B., Smith, A.L., 2013. Conceptual 690 domain of the matrix in fragmented landscapes. Trends Ecol. Evol. 28, 605–613. 691 https://doi.org/10.1016/J.TREE.2013.06.010 692 Erwin, R.M., Hatfield, J.S., Wilmers, T.J., 1995. The value and vulnerability of small estuarine 693 islands for conserving metapopulations of breeding waterbirds. Biol. Conserv. 71, 187–191. 694 https://doi.org/10.1016/0006-3207(94)00045-R 695 Estades, C.F., 2001. The effect of breeding-habitat patch size on bird population density. Landsc. 696 Ecol. 16, 161–173. https://doi.org/10.1023/A:1011197432467 697 Evans, M.C., 2016. Deforestation in Australia: drivers, trends and policy responses. Pacific 698 Conserv. Biol. 22, 130–150. https://doi.org/https://doi.org/10.1071/PC15052 699 Fahrig, L., 2013. Rethinking patch size and isolation effects: the habitat amount hypothesis. J. 700 Biogeogr. 40, 1649–1663. https://doi.org/10.1111/jbi.12130 701 Fischer, J., Lindenmayer, D.B., 2002. Small patches can be valuable for biodiversity conservation: 702 two case studies on birds in southeastern Australia. Biol. Conserv. 106, 129–136. 703 Flaspohler, D.J., Giardina, C.P., Asner, G.P., Hart, P., Price, J., Lyons, C.K., Castaneda, X., 2010. 704 Long-term effects of fragmentation and fragment properties on bird species richness in 705 Hawaiian forests. Biol. Conserv. 143, 280–288. 706 https://doi.org/http://dx.doi.org/10.1016/j.biocon.2009.10.009 707 Fletcher, Jr., R.J., Ries, L., Battin, J., Chalfoun, A.D., 2007. The role of habitat area and edge in 708 fragmented landscapes: definitively distinct or inevitably intertwined?This review is one of a 709 series dealing with some aspects of the impact of habitat fragmentation on animals and plants. 710 This series is one of several virtual symposia focussing on ecological topics that will be 711 published in the Journal from time to time. Can. J. Zool. 85, 1017–1030. 712 https://doi.org/10.1139/Z07-100 713 Ford, H.A., Walters, J.R., Cooper, C.B., Debus, S.J.S., Doerr, V.A.J., 2009. Extinction debt or 714 habitat change?–Ongoing losses of woodland birds in north-eastern New South Wales, 715 Australia. Biol. Conserv. 142, 3182–3190. 716 Freeman, A.N.D., Freeman, A.B., Burchill, S., 2009. Bird use of revegetated sites along a creek 717 connecting rainforest remnants. Emu 109:331–338. http://dx.doi.org/10.1071/MU09089 718 Freudenberger, D., 2001. “Bush for the birds: Biodiversity enhancement guidelines for the 719 Saltshaker Project, Boorowa, NSW,” Consultancy report to Greening Australia ACT & SE 720 NSW, Inc. CSIRO Sustainable Ecosystems, Canberra, Australia. 721 Gibbons, P., Boak, M., 2002. The value of paddock trees for regional conservation in an 722 agricultural landscape. Ecol. Manag. Restor. 3, 205–210. https://doi.org/10.1046/j.1442- 42

723 8903.2002.00114.x 724 Gill, F., Donsker, D., 2018. IOC World Bird List [WWW Document]. URL 725 http://www.worldbirdnames.org/ioc-lists/master-list-2/ (accessed 8.14.18). 726 Goslee, S.C., Urban, D.L., 2007. The ecodist package for dissimilarity-based analysis of ecological 727 data. J. Stat. Softw. 22, 1–19. 728 Hadley, A.S., Frey, S.J.K., Robinson, W.D., Betts, M.G., 2018. Forest fragmentation and loss 729 reduce richness, availability, and specialization in tropical hummingbird communities. 730 Biotropica 50, 74–83. https://doi.org/10.1111/btp.12487 731 Helzer, C.J., Jelinski, D.E., 1999. The relative importance of patch area and perimeter-ratio to 732 grassland breeding birds. Ecol. Appl. 9, 1448–1458. https://doi.org/10.1890/1051- 733 0761(1999)009[1448:TRIOPA]2.0.CO;2 734 Herkert, J.R., Reinking, D.L., Wiedenfeld, D.A., Winter, M., Zimmerman, J.L., Jensen, W.E., 735 Finck, E.J., Koford, R.R., Wolfe, D.H., Sherrod, S.K., 2003. Effects of prairie fragmentation 736 on the nest success of breeding birds in the midcontinental United States. Conserv. Biol. 17, 737 587–594. 738 Hinsley, S.A., Bellamy, P.E., Newton, I., Sparks, T.H., 1995. Habitat and landscape factors 739 influencing the presence of individual breeding bird species in woodland fragments. J. Avian 740 Biol. 26, 94–104. https://doi.org/10.2307/3677057 741 Hinsley, S.A., Hill, R.A., Bellamy, P.E., Harrison, N.M., Speakman, J.R., Wilson, A.K., Ferns, 742 P.N., 2008. Effects of structural and functional habitat gaps on breeding woodland birds: 743 working harder for less. Landsc. Ecol. 23, 615–626. 744 Hoover, J.P., Brittingham, M.C., Goodrich, L.J., 1995. Effects of forest patch size on nesting 745 success of Wood Thrushes. Auk 112, 146–155. https://doi.org/10.2307/4088774 746 Hui, F.K.C., 2016. boral - Bayesian Ordination and Regression Analysis of Multivariate Abundance 747 Data in r. Methods Ecol. Evol. 7, 744–750. https://doi.org/10.1111/2041-210X.12514 748 Ikin, K., Yong, D.L., Lindenmayer, D.B., 2016. Effectiveness of woodland birds as taxonomic 749 surrogates in conservation planning for biodiversity on farms. Biol. Conserv. 204, Part, 411– 750 416. https://doi.org/http://dx.doi.org/10.1016/j.biocon.2016.11.010 751 Kahle D., Wickham, H., 2013. ggmap: Spatial Visualization with ggplot2. R J 5:144–161 752 Kavanagh, R.P., Stanton, M.A., Herring, M.W., 2007. Eucalypt plantings on farms benefit 753 woodland birds in south‐eastern Australia. Austral Ecol. 32, 635–650. 754 Kendal, D., Zeeman, B.J., Ikin, K., Lunt, I.D., McDonnell, M.J., Farrar, A., Pearce, L.M., Morgan, 755 J.W., 2017. The importance of small urban reserves for plant conservation. Biol. Conserv. 213, 756 146–153. https://doi.org/10.1016/J.BIOCON.2017.07.007 757 King, D.I., Chandler, R.B., Collins, J.M., Petersen, W.R., Lautzenheiser, T.E., 2009. Effects of 758 width, edge and habitat on the abundance and nesting success of scrub–shrub birds in 759 powerline corridors. Biol. Conserv. 142, 2672–2680. 760 https://doi.org/http://dx.doi.org/10.1016/j.biocon.2009.06.016 761 Kinross, C.M., 2004. Avian use of farm habitats, including windbreaks, on the New South Wales 762 Tablelands. Pacific Conserv. Biol. 10, 180–192. 763 Kinross, C.M., Nicol, H., 2008. Responses of birds to the characteristics of farm windbreaks in 764 central New South Wales, Australia. Emu 108:139–152. http://dx.doi.org/10.1071/MU06024 765 Le Roux, D.S., Ikin, K., Lindenmayer, D.B., Manning, A.D., Gibbons, P., 2015. Single large or 766 several small? Applying biogeographic principles to tree-level conservation and biodiversity 767 offsets. Biol. Conserv. 191, 558–566. 768 https://doi.org/http://dx.doi.org/10.1016/j.biocon.2015.08.011 769 Lehnen, S.E., Rodewald, A.D., 2009. Investigating area-sensitivity in shrubland birds: Responses to 770 patch size in a forested landscape. For. Ecol. Manage. 257, 2308–2316. 771 https://doi.org/http://dx.doi.org/10.1016/j.foreco.2009.03.008 772 Lindenmayer, D.B., Blanchard, W., Crane, M., Michael, D., Florance, D., 2018a. Size or quality. 773 What matters in vegetation restoration for bird biodiversity in endangered temperate 774 woodlands? Austral Ecol. https://doi.org/10.1111/aec.12622 43

775 Lindenmayer D.B., Blanchard, W., Crane, M., Michael, D., Sato, C., 2018b. Biodiversity benefits 776 of vegetation restoration are undermined by livestock grazing. Restor Ecol 26, 1157–1164. 777 https://doi.org/10.1111/rec.12676 778 Lindenmayer, D.B., Cunningham, R.B., Crane, M.J., Michael, D.R., Montague-Drake, R.M., 2007. 779 Farmland bird responses to intersecting replanted areas. Landsc. Ecol. 22, 1555–1562. 780 https://doi.org/10.1007/s10980-007-9156-9 781 Lindenmayer, D.B., Knight, E.J., Crane, M.J., Montague-Drake, R.M., Michael, D.R., MacGregor, 782 C.I., 2010. What makes an effective restoration planting for woodland birds? Biol. Conserv. 783 143, 289–301. 784 Lindenmayer, D.B., Lane, P.W., Barton, P.S., Crane, M.J., Ikin, K., Michael, D.R., Okada, S., 2016. 785 Long-term bird colonization and turnover in restored woodlands. Biodivers Conserv 25:1587– 786 1603. https://doi.org/10.1007/s10531-016-1140-8 787 Lindenmayer, D.B., Northrop-Mackie, A.R., Montague-Drake, R.M., Crane, M.J., Michael, D.R., 788 Okada, S., Gibbons, P., 2012. Not all kinds of revegetation are created equal: revegetation type 789 influences bird assemblages in threatened Australian woodland ecosystems. PLoS One 7, 790 e34527. 791 Lindenmayer, D.B., Willinck, E., Crane, M.J., Michael, D.R., Okada, S., Cumming, C., Durant, K., 792 Frankenberg, J., 2013. Murray Catchment habitat restoration: Lessons from landscape-level 793 research and monitoring. Ecol. Manag. Restor. 14, 80–92. https://doi.org/10.1111/emr.12051 794 Loyn, R.H., McNabb, E.G., Macak, P., Noble, P., 2007. Eucalypt plantations as habitat for birds on 795 previously cleared farmland in south-eastern Australia. Biol. Conserv. 137, 533–548. 796 Luck, G.W., 2003. Differences in the reproductive success and survival of the rufous treecreeper 797 (Climacteris rufa) between a fragmented and unfragmented landscape. Biol. Conserv. 109, 1– 798 14. 799 Mac Nally, R., 2007. Consensus weightings of evidence for inferring breeding success in broad‐ 800 scale bird studies. Austral Ecol. 32, 479–484. 801 Mac Nally, R., De Vries, L., Thomson, J.R., 2010. Are replanted floodplain forests in southeastern 802 Australia providing bird biodiversity benefits? Restor. Ecol. 18, 85–94. 803 MacArthur, R.H., Wilson, E.O., 1967. The Theory of Island Biogeography. Princeton University 804 Press. 805 Major, R.E., Gowing, G., 2001. Survival of red-capped robins (Petroica goodenovii) in woodland 806 remnants of central western New South Wales, Australia. Wildl. Res. 28, 565–571. 807 https://doi.org/http://dx.doi.org/10.1071/WR01040 808 Major, R.E., Christie, F.J., Gowing, G., 2001. Influence of remnant and landscape attributes on 809 Australian woodland bird communities. Biol Conserv 102:47–66. 810 Manning, A.D., Fischer, J., Lindenmayer, D.B., 2006. Scattered trees are keystone structures – 811 Implications for conservation. Biol. Conserv. 132, 311–321. 812 https://doi.org/http://dx.doi.org/10.1016/j.biocon.2006.04.023 813 Maron, M., Grey, M.J., Catterall, C.P., Major, R.E., Oliver, D.L., Clarke, M.F., Loyn, R.H., Mac 814 Nally, R., Davidson, I., Thomson, J.R., 2013. Avifaunal disarray due to a single despotic 815 species. Divers. Distrib. 19, 1468–1479. https://doi.org/10.1111/ddi.12128 816 Martin, W.K., Eldridge, D., Murray, P.A., 2011. Bird assemblages in remnant and revegetated 817 habitats in an extensively cleared landscape, Wagga Wagga, New South Wales. Pacific 818 Conserv. Biol. 17, 110–120. 819 Martin, W.K., Eyears-Chaddock, M., Wilson, B.R., Lemon, J., 2004. The value of habitat 820 reconstruction to birds at Gunnedah, New South Wales. Emu 104, 177–189. 821 Maxwell, S.L., Fuller, R.A., Brooks, T.M., Watson, J.E.M., 2016. Biodiversity: The ravages of 822 guns, nets and bulldozers. Nature 536:143–145. doi: 10.1038/536143a 823 Montague-Drake, R.M., Lindenmayer, D.B., Cunningham, R.B., 2009. Factors affecting site 824 occupancy by woodland bird species of conservation concern. Biol. Conserv. 142, 2896–2903. 825 Morcombe, M.K., 2003. Field guide to Australian birds. Steve Parish Publishing. 826 Munro, N.T., Fischer, J., Barrett, G.W., Wood, J.T., Leavesley, A., Lindenmayer, D.B., 2011. 44

827 Bird’s response to revegetation of different structure and floristics—Are “restoration 828 plantings” restoring bird communities? Restor. Ecol. 19, 223–235. 829 NSW Environment & Heritage, 2018. Threatened Species found in Inland Slopes IBRA sub-region. 830 https://www.environment.nsw.gov.au/threatenedSpeciesApp/cmaSearchResults.aspx?SubCma 831 Id=344 (accessed 8.14.18). 832 Okada, S., Lindenmayer, D.B., Wood, J.T., Crane, M.J., Pierson, J.C., 2017. How does a 833 transforming landscape influence bird breeding success? Landsc. Ecol. 32, 1039–1048. 834 https://doi.org/10.1007/s10980-017-0507-x 835 Parsons, B.C., Gosper, C.R., 2011. Contemporary fire regimes in a fragmented and an 836 unfragmented landscape: implications for vegetation structure and persistence of the fire- 837 sensitive malleefowl. Int J Wildl Fire 20:184. https://doi.org/10.1071/WF09099 838 Pizzey, G., Knight, F., 1997. The Graham Pizzey and Frank Knight Field Guide to the Birds of 839 Australia. Angus & Robertson, Sydney, New South Wales. 840 Radford, J.Q., Bennett, A.F., 2007. The relative importance of landscape properties for woodland 841 birds in agricultural environments. J. Appl. Ecol. 44, 737–747. 842 Rayner, L., Lindenmayer, D.B., Gibbons, P., Manning, A.D., 2014. Evaluating empirical evidence 843 for decline in temperate woodland birds: A nationally threatened assemblage of species. Biol. 844 Conserv. 171, 145–155. https://doi.org/http://dx.doi.org/10.1016/j.biocon.2014.01.029 845 Ribeiro, Jr, P.J., Diggle, P.J., 2016. geoR: Analysis of Geostatistical Data. R package version 1.7- 846 5.2. https://cran.r-project.org/package=geoR 847 Root, R.B., 1973. Organization of a Plant-Arthropod Association in Simple and Diverse Habitats: 848 The Fauna of Collards (Brassica oleracea). Ecol. Monogr. 43, 95–124. 849 https://doi.org/10.2307/1942161 850 Rosenzweig, M.L., 1995. Species diversity in space and time. Cambridge University Press. 851 Ruiz-Jaen, M.C., Aide, T.M., 2005. Restoration success: How is it being measured? Restor. Ecol. 852 13, 569–577. https://doi.org/10.1111/j.1526-100X.2005.00072.x 853 Schippers, P., Grashof-Bokdam, C.J., Verboom, J., Baveco, J.M., Jochem, R., Meeuwsen, H.A.M., 854 Van Adrichem, M.H.C., 2009. Sacrificing patches for linear habitat elements enhances 855 metapopulation performance of woodland birds in fragmented landscapes. Landsc. Ecol. 24, 856 1123–1133. 857 Selwood, K., Mac Nally, R., Thomson, J.R., 2009. Native bird breeding in a chronosequence of 858 revegetated sites. Oecologia 159, 435–446. 859 Shanahan, D.F., Miller, C., Possingham, H.P., Fuller, R.A., 2011. The influence of patch area and 860 connectivity on avian communities in urban revegetation. Biol. Conserv. 144, 722–729. 861 https://doi.org/http://dx.doi.org/10.1016/j.biocon.2010.10.014 862 Silcocks, A., Tzaros, C., Weston, M., Olsen, P., 2005. An interim guild classification for woodland 863 and grassland birds in Australia. Birds Australia Supplementary Report to State of the 864 Environment Report 2006, Carlton. 865 Smith, F.P., 2008. Who’s planting what, where and why – and who’s paying?: An analysis of 866 farmland revegetation in the central wheatbelt of Western Australia. Landsc. Urban Plan. 86, 867 66–78. https://doi.org/http://dx.doi.org/10.1016/j.landurbplan.2007.12.008 868 Stephens, S.E., Koons, D.N., Rotella, J.J., Willey, D.W., 2004. Effects of habitat fragmentation on 869 avian nesting success: a review of the evidence at multiple spatial scales. Biol. Conserv. 115, 870 101–110. 871 Thiele, K.R., Prober, S.M., 2000. Reserve concepts and conceptual reserves: options for the 872 protection of fragmented ecosystems, in: Hobbs, R.J., Yates, C.J. (Eds.), Temperate Eucalypt 873 Woodlands in Australia: Biology, Conservation, Management and Restoration. Surrey Beatty 874 and Sons, Chipping Norton, pp. 351–358. 875 Tilman, D., Clark, M., Williams, D.R., Kimmel, K., Polasky, S., Packer, C., 2017. Future threats to 876 biodiversity and pathways to their prevention. Nature 546, 73–81. 877 https://doi.org/10.1038/nature22900 878 Tscharntke, T., Tylianakis, J.M., Rand, T.A., Didham, R.K., Fahrig, L., Batáry, P., Bengtsson, J., 45

879 Clough, Y., Crist, T.O., Dormann, C.F., Ewers, R.M., Fründ, J., Holt, R.D., Holzschuh, A., 880 Klein, A.M., Kleijn, D., Kremen, C., Landis, D.A., Laurance, W., Lindenmayer, D.B., 881 Scherber, C., Sodhi, N., Steffan-Dewenter, I., Thies, C., van der Putten, W.H., Westphal, C., 882 2012. Landscape moderation of biodiversity patterns and processes - eight hypotheses. Biol. 883 Rev. 87, 661–685. https://doi.org/10.1111/j.1469-185X.2011.00216.x 884 Vesk, P.A., Nolan, R., Thomson, J.R., Dorrough, J.W., Mac Nally, R., 2008. Time lags in provision 885 of habitat resources through revegetation. Biol Conserv 141:174–186. 886 http://dx.doi.org/10.1016/j.biocon.2007.09.010 887 Watson, J.E.M., Freudenberger, D., Paull, D., 2001. An assessment of the focal-species approach 888 for conserving birds in variegated landscapes in southeastern Australia. Conserv. Biol. 15, 889 1364–1373. 890 Watson, J.E.M., Watson, A., Paull, D., Freudenberger, D., 2003. Woodland fragmentation is 891 causing the decline of species and functional groups of birds in southeastern Australia. Pacific 892 Conserv. Biol. 8, 261–270. 893 Watson, J.E.M., Whittaker, R.J., Freudenberger, D., 2005. Bird community responses to habitat 894 fragmentation: how consistent are they across landscapes? J. Biogeogr. 32, 1353–1370. 895 Westphal, M.I., Field, S.A., Possingham, H.P., 2007. Optimizing landscape configuration: a case 896 study of woodland birds in the Mount Lofty Ranges, South Australia. Landsc. Urban Plan. 81, 897 56–66. 898 Wintle, B.A., Kujala, H., Whitehead, A., Cameron, A., Veloz, S., Kukkala, A., Moilanen, A., 899 Gordon, A., Lentini, P.E., Cadenhead, N.C.R., Bekessey, S.A., 2019. Global synthesis of 900 conservation studies reveals the importance of small habitat patches for biodiversity. Proc Natl 901 Acad Sci U S A 116:909–914. https://doi.org/10.1073/pnas.1813051115 902 Woolhouse, M.E.J., 1983. The theory and practice of the species-area effect, applied to the breeding 903 birds of British Woods. Biol. Conserv. 27, 315–332. https://doi.org/10.1016/0006- 904 3207(83)90089-7 905 Zanette, L., 2001. Indicators of habitat quality and the reproductive output of a forest songbird in 906 small and large fragments. J. Avian Biol. 32, 38–46. https://doi.org/10.1034/j.1600- 907 048X.2001.320106.x 908 Zanette, L., Doyle, P., Trémont, S.M., 2000. Food shortage in small fragments: Evidence from an 909 area-sensitive passerine. Ecology 81, 1654–1666. https://doi.org/10.1890/0012- 910 9658(2000)081[1654:FSISFE]2.0.CO;2 911 Zanette, L., Jenkins, B., 2000. Nesting success and nest predators in forest fragments: A study using 912 real and artificial nests. Auk 117, 445–454. https://doi.org/10.1642/0004- 913 8038(2000)117[0445:NSANPI]2.0.CO;2 914

46